[gmx-users] GROMACS and CNT/graphene (probably for the 1000th time)
Hi All, I'm sure you've seen this many times before, but I think it might help those who have never done this before, are new to Gromacs, or, like me, haven't done anything relatively advanced with it in a while. So, here goes. My intermediate goal here was a solvated sheet of graphene, so I followed http://www.gromacs.org/Documentation/How-tos/Carbon_Nanotube under Andrea Minoia's tutorial pretty much to the word (there is a link with the exact file content). Everything prior to solvating the system worked without a hitch. To solvate the system, I used genbox, making it also update the topology file (created with g_x2top). This is when grompp started screaming about defaults duplicates, and, as I tried removing the #include to cnt_oplsaa.ff/forcefield.itp from the topology file and just put all parameters explicitly in there, there was a bunch of other problems. Without going into detail of the latter, I don't think I know what I am doing at this point. Can you please help with this task in a simple robust step-by-step way? In fact, I'd like to write up an updated tutorial based on our discussion, because, among other things, Andrea's parameters are taken from oplsaa on aromatics, which can and should be further refined. All input is greatly appreciated and I'd especially love to hear from the Gromacs superhero Justin Lemkul. :) Thanks! Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] GROMACS and CNT/graphene (probably for the 1000th time)
You guys have thought of everything, haven't you... :) Thanks. JL On 3/23/15 9:09 PM, Alex wrote: Great, thanks. One other question, just sort of jumping way ahead. Let's say I had a denaturated protein chain I wanted to drag in a simulation of translocation. Is there anything in Gromacs for that, aside from fake charges and external electric fields, or messing with the code itself? JL Use the pull code. JL -Justin JL -- JL == JL Justin A. Lemkul, Ph.D. JL Ruth L. Kirschstein NRSA Postdoctoral Fellow JL Department of Pharmaceutical Sciences JL School of Pharmacy JL Health Sciences Facility II, Room 629 JL University of Maryland, Baltimore JL 20 Penn St. JL Baltimore, MD 21201 JL jalem...@outerbanks.umaryland.edu | (410) 706-7441 JL http://mackerell.umaryland.edu/~jalemkul JL == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
[gmx-users] OPLS/AA support for DNA
I think I just googled my head off trying to find out if oplsaa supports DNA chains, and I just can't find a definitive answer. In a 2008 post, Justin suggested that AMBER is a better choice for DNA, but more recent posts indicate that people casually use oplsaa with DNA (at least that's my understanding). There are some reasons I'd like to stick with oplsaa with my project. Not set in stone, but highly desirable. Any ideas? Thanks, Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Setting custom atoms in FF
I am not sure if your answer was a resounding yes. I am talking about the label on the left in PDB. Let us say I want to name my atoms TEST1 and TEST2 and set up my PDB as ATOM 1 TEST1 CNT A 1 xxx yyy zzz 1.00 0.00 ATOM 2 TEST2 CNT A 1 xxx yyy zzz 1.00 0.00 and then in atomname2type.a2t I put TEST1 opls_9960 12.011 3C 0.142 C 0.142 C 0.142 TEST2 opls_1001 0.112.011 3C 0.142 C 0.142 C 0.149 will the resulting topology contain the distinction despite the issue with bond lengths? If not, what would be the correct syntax to do that? I just want to define my atoms in a way that supplements (actually, precedes) what x2top does when the atoms are not specifically labeled. A tedious solution in my case, but definitely viable, if this works. Thank you, Alex JL On 3/29/15 7:53 PM, Alex wrote: Another question about x2top: in the previous discussion, the atom type was determined from its environment and the element type. What if I set up my PDB with atom labels (i think that's the third field from the left in the PDB format), say CJ1 and CJ2 (instead of just C), while the force field contains CJ1 opls_9960 12.011 3C 0.142 C 0.142 C 0.142 CJ2 opls_1001 0.112.011 3C 0.142 C 0.142 C 0.149 Will then the right distinction be made? I'm asking before trying it, because it is quite tedious for me to do the labeling. The issue is way beyond carbons, I just discovered that a bunch of different atoms isn't set up correctly. JL The element field can be up to 16 characters long, per the code (overkill, but JL it indicates that there should be some flexibility). JL -Justin JL -- JL == JL Justin A. Lemkul, Ph.D. JL Ruth L. Kirschstein NRSA Postdoctoral Fellow JL Department of Pharmaceutical Sciences JL School of Pharmacy JL Health Sciences Facility II, Room 629 JL University of Maryland, Baltimore JL 20 Penn St. JL Baltimore, MD 21201 JL jalem...@outerbanks.umaryland.edu | (410) 706-7441 JL http://mackerell.umaryland.edu/~jalemkul JL == -- Best regards, Alexmailto:nedoma...@gmail.com -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Density of pure tetrolic acid does not agree with experimental data
Nathan, The experimental data ideally corresponds to an infinitely long, well-converged MD simulation at the corresponding set of temperature and pressure, not after EM and most certainly not after just filling a box, where spacing is according to a preset cut-off and not due to interatomic interactions. Alex NKH Hello Gromacs users, NKH I tried to create a 6.5 nm cubic box of tetrolic acid (otherwise NKH known as 2-butynoic acid), but the density is too low. According NKH to this, http://www.chemspider.com/Chemical-Structure.61810.html, NKH the density should be about 0.964 NKH g/mL. Given the molecular weight of tetrolic acid, this NKH corresponds to right around 1900 molecules in the 6.5 nm box I NKH prepared. Unfortunately, when I attempted to fill the box, NKH Gromacs only found room for 1676 molecules, resulting in NKH a density of 0.852 g/mL. I know that models are not 100% NKH accurate, but I'm worried about a difference that big. Should I NKH expect the system to condense when I minimize, and just shrink NKH the box afterwards? Or do you think there is NKH another problem? NKH Thanks for your help, NKH Nathan -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Setting custom atoms in FF
Another question about x2top. The directives I am putting in ffbonded explicitly state that my angle is of type 2 (G96). Why is it reverting to its default value of 1 in the output topology? It wasn't a problem before, when I was just testing a graphene sheet. Now, it's turning into a mess... I know x2top has limited intelligence, but how is this correct program behavior? Also, is there any way to tell x2top to copy all bond and angle parameters from the ff data into the topology the way pdb2gmx does? Thanks, Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Setting custom atoms in FF
Building things is not necessary. I just wasn't sure what particular field in PDB you were talking about. Some indication that it is supposed to work is enough. ;) Thanks, Alex JL I don't have the time or desire to build something, test it, and give you an JL absolute answer :) I can only tell you what the code says, which I can do quickly. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Setting custom atoms in FF
Wha, that's a huge blow, and it explains a bunch of other issues. I was thinking the difference had to be around 2-3% and set my pdb files very precisely. I actually think that either the tolerance is even larger, or there's something on top of that in the code. Setting CJ opls_9960 12.011 3C 0.10 C 0.10 C 0.10 ;bulk C still assigns #996 to bulk graphene carbons, despite the difference of about 42%. Anyway, this is nearly catastrophic for what I'm trying to do. From what I understand, the tolerance cannot be set in the x2top command itself... Oy. If you know it off the top of your head, could you please point me to the variable name in g_x2top.c (if it's there)? I'm using a precompiled version right now, but I guess I will have to mess with the code on our cluster. Allowing this tolerance to be set either as percentage, or in distance units would be extremely useful. One could always use the default value if no setting is entered. Could the last point be relayed to the developers? Sounds like a huge necessity to me, to be honest. Thank you, Alex JL On 3/29/15 4:07 PM, Alex wrote: Hi all, I am messing with the OPLS/AA forcefield. The idea is to functionalize graphene edge atoms beyond passivation with hydrogen. At the functionalization region, there is a C atom with two C-C bonds (from graphene), while the third bond is formed with the functional group. The particular atom from the functional group intended to bond with that C bond is also carbon. When setting up the system, in atomname2type.n2t I have: CJ opls_9960 12.011 3C 0.142 C 0.142 C 0.142 ; bulk graphene C but also want to introduce CJ opls_1001 0.112.011 3C 0.142 C 0.142 C 0.149 ; edge C Both 996 and 1001 are defined in the ffnononbonded.itp, these are custom, but essentially copies from oplsaa entries for carbon. Note the third bond that's longer (which is also the case in my input PDB). The significant difference is of course the charge there, as I want a globally neutral system. Here is the problem: x2top ignores the second statement from above and the functionalized edge C is assigned #996. Aside from manually modifying the resulting topology, can we make this distinction stick? JL g_x2top (and other programs like pdb2gmx) use a default ± 10% distance tolerance JL for deciding if something is within bonding distance. The distinction between JL 0.149 and 0.142 is about 5%, so the two lines are effectively indistinguishable. JL The most straightforward solution would be to change the default tolerance in JL the code, if your distances are going to be so close in magnitude. JL -Justin JL -- JL == JL Justin A. Lemkul, Ph.D. JL Ruth L. Kirschstein NRSA Postdoctoral Fellow JL Department of Pharmaceutical Sciences JL School of Pharmacy JL Health Sciences Facility II, Room 629 JL University of Maryland, Baltimore JL 20 Penn St. JL Baltimore, MD 21201 JL jalem...@outerbanks.umaryland.edu | (410) 706-7441 JL http://mackerell.umaryland.edu/~jalemkul JL == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
[gmx-users] Setting custom atoms in FF
Hi all, I am messing with the OPLS/AA forcefield. The idea is to functionalize graphene edge atoms beyond passivation with hydrogen. At the functionalization region, there is a C atom with two C-C bonds (from graphene), while the third bond is formed with the functional group. The particular atom from the functional group intended to bond with that C bond is also carbon. When setting up the system, in atomname2type.n2t I have: CJ opls_9960 12.011 3C 0.142 C 0.142 C 0.142 ; bulk graphene C but also want to introduce CJ opls_1001 0.112.011 3C 0.142 C 0.142 C 0.149 ; edge C Both 996 and 1001 are defined in the ffnononbonded.itp, these are custom, but essentially copies from oplsaa entries for carbon. Note the third bond that's longer (which is also the case in my input PDB). The significant difference is of course the charge there, as I want a globally neutral system. Here is the problem: x2top ignores the second statement from above and the functionalized edge C is assigned #996. Aside from manually modifying the resulting topology, can we make this distinction stick? Thanks, Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Setting custom atoms in FF
Yeah, I know there's that disclaimer. I wonder if this http://software-lisc.fbk.eu/obgmx/ could be useful. Any comment on this UFF? First time I'm hearing about it, to be honest. The system I have in mind isn't much beyond CNTs and nucleotide chains. Well, and water. Thanks. Alex JL g_x2top also doesn't purport to be terribly smart. It may be better to find JL another means to generate your topology if there are going to be complex systems JL and subtle differences. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Setting custom atoms in FF
Another question about x2top: in the previous discussion, the atom type was determined from its environment and the element type. What if I set up my PDB with atom labels (i think that's the third field from the left in the PDB format), say CJ1 and CJ2 (instead of just C), while the force field contains CJ1 opls_9960 12.011 3C 0.142 C 0.142 C 0.142 CJ2 opls_1001 0.112.011 3C 0.142 C 0.142 C 0.149 Will then the right distinction be made? I'm asking before trying it, because it is quite tedious for me to do the labeling. The issue is way beyond carbons, I just discovered that a bunch of different atoms isn't set up correctly. Thanks, Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] OPLS/AA support for DNA
Well, just because I am asking about Gromacs, doesn't mean I don't know about them interactions, especially the ones as mathematically simple as those in Gromacs... Having said that, I agree that the chain flexibility is greatly affected by bonded parameter issues. However, the particular simulation I have in mind (translocation of a near-linear ssdna chain periodic in length) is not going to be drastically affected in this case, i.e. the order of magnitude of the desired phenomena wouldn't change. There will be some effects, of course, but I will be very suprised if it's drastic. I just come from solid-state systems and wish I had experience with Gromacs that's beyond editconf-genbox-pdb2gmx-grompp. Come to think of it, you'd set up my system in about an hour. Care to collaborate? :))) Alex JL On 3/26/15 6:10 PM, Alex wrote: Yeah, I agree with Justin, this is pretty half-baked. On the other hand, I am not after conformational changes, ds-DNA simulations, or things of that nature. The things I am interested in would be mostly affected by the partial charges and vdW parameters, and all of those seem to be from standard oplsaa. JL Everything in force fields is interrelated, and nucleic acids are particularly JL prone to deficiencies in bonded parameters, as the backbone dynamics are highly JL correlated. There's a reason why nucleic acid force fields (especially in the JL AMBER and CHARMM families) have undergone continual development for the better JL part of 20 years. JL See, for instance, dx.doi.org/10.1021/jz500557y JL -Justin JL -- JL == JL Justin A. Lemkul, Ph.D. JL Ruth L. Kirschstein NRSA Postdoctoral Fellow JL Department of Pharmaceutical Sciences JL School of Pharmacy JL Health Sciences Facility II, Room 629 JL University of Maryland, Baltimore JL 20 Penn St. JL Baltimore, MD 21201 JL jalem...@outerbanks.umaryland.edu | (410) 706-7441 JL http://mackerell.umaryland.edu/~jalemkul JL == -- Best regards, Alexmailto:nedoma...@gmail.com -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] OPLS/AA support for DNA
Yeah, I agree with Justin, this is pretty half-baked. On the other hand, I am not after conformational changes, ds-DNA simulations, or things of that nature. The things I am interested in would be mostly affected by the partial charges and vdW parameters, and all of those seem to be from standard oplsaa. Alex On Thu, Mar 26, 2015 at 3:56 PM, Justin Lemkul jalem...@vt.edu wrote: On 3/26/15 8:30 AM, Smith, Micholas D. wrote: There was some work on OPLS-AA/L on nucleic acids back in version 3 of GROMACS. The group that did the work has a website where they discuss the nucleic parameters located at: http://rnp-group.genebee.msu.su/3d/ff.htm There are likely newer parameters out there for opls, but I haven't looked very hard for them. Anyone heard of anything peer-reviewed? I know I'm not convinced by 2-ns simulations :) -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/ Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] GROMACS and CNT/graphene (probably for the 1000th time)
Hey Justin, Thanks. I am not entirely sure I understand your suggestion, because this will clearly work if nothing else is added to the system (it seems to be equivalent to that CNT guide). The problems start after I add water, i.e. duplicate [ defaults ] from grompp and such. In the meantime, I will try the earlier suggestion and add the bonded definitions on top of what was suggested. Alex JL For this, the cleanest approach is to create a custom force field, and its JL contents should be rather simple - bonded and nonbonded parameters for some C JL type (for the CNT) and whatever solvent type(s) you need. It's actually a very JL bare-bones force field. The proposal before about using values from OPLS-AA is JL a good guide; these are the things you need to define. In reality, this could JL all be in one file: JL [ defaults ] JL ...whatever... JL [ atomtypes ] JL ...whatever... JL [ nonbond_params ] JL (if needed) JL [ bondtypes ], [ angletypes ], [ dihedraltypes ], ... JL Then create the .n2t file for creating the CNT. That part is the easiest of all JL of it since everything is (should be) equally spaced. :) JL -Justin -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] GROMACS and CNT/graphene (probably for the 1000th time)
Hi Bogdan,
I agree with your statement: Gromacs is an entirely inappropriate choice for a solid-state system, the lack of proper potential functional forms ("forcefields" in biomolecular slang) is chief among them. However, Gromacs shines in biomolecular simulations, and what I have in mind is one of those, as I'm sure many other researchers have. It just happens to involve graphene. ;)
So, it may be worthwhile to gain some level of flexibility to define our own bonded parameters, if covalent bond creation/breake is not on the table.
Thanks,
Alex
I was in your situation about some years ago and couldn't find any way around generating topologies on my own, without using the GROMACS tools. As you could already see most GROMACS tools are not suitable for graphene/CNT type of systems. Topologies are well documented and you gain enormously in flexibility by tailoring the process to your needs.
Good luck,
Bogdan
--
Gromacs Users mailing list
* Please search the archive at
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!
* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a
mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] GROMACS and CNT/graphene (probably for the 1000th time)
Okay, I am learning here. Here's what I did: 1. Copied the entire oplsaa folder in the local directory. 2. Added the following to ffnonbonded.itp: opls_995 CJ 612.01100 0.000 A3.4e-01 3.61200e-01 opls_996 CJ 612.01100 0.000 A3.4e-01 3.61200e-01 opls_997 CJ 612.01100 0.000 A3.4e-01 3.61200e-01 3. Added the following to ffbonded.itp under dihedral types: CJ CJ CJ CJ 3 17.30770 0.0 -17.30770 0.0 0.0 0.0 ; From opt. Brenner under angle types: CJ CJ CJ 2 120.000776.923 ; G96 angle, param from opt. Brenner under bond types: CJCJ 10.14200 420420.0 ; from opt. Brenner 698.129 N/m 4. Added the following to atomname2type.n2t: CJ opls_9950 12.011 2C 0.142 C 0.142 CJ opls_9960 12.011 3C 0.142 C 0.142 C 0.142 CJ opls_9970 12.011 1C 0.142 After all of this g_x2top with -noparam does almost everything right (the angle type needs to be manually changed to 2, although it is explicitly 2 above). But whatever, I can run the in vacuo dynamic simulation, and the sheet behaves beautifully and respects my PBC (the box has precise in-plane dimensions). I have solvated my sheet with quite a bit of SPC water (~32K molecules), about to test some energy minimization followed by a dynamic simulation. Does the above sound reasonable? Thanks, Alex p.s. The structure alone was working right from that tutorial, it's the solvation part that sucked. JL If you're introducing a new [defaults] directive, then whatever you're doing is JL breaking the required force field format. Chapter 5 is your best friend here. JL You say you want some custom parameters and that you're having issues with JL implementing new parameter into existing force fields, so I'm telling you that JL the best approach (in my mind) is to do something simple. Something that you JL can write in a few lines (for a CNT you have a single C type and I can count the JL bonded parameters on one hand). Make sure that works. Then worry about making JL it more complicated. But at that point, you're just adding more [atomtypes] and JL [bond|angle|dihedraltypes], nothing more. You only ever have one [defaults] - JL these are the master instructions for the shape of the functional form and how JL the force field is interpreted. JL Canary in the mineshaft - get the CNT working alone. Make sure you understand JL everything about that. Because at that point, you've done most of the hard work JL and everything is just an addition to it. JL -Justin JL -- JL == JL Justin A. Lemkul, Ph.D. JL Ruth L. Kirschstein NRSA Postdoctoral Fellow JL Department of Pharmaceutical Sciences JL School of Pharmacy JL Health Sciences Facility II, Room 629 JL University of Maryland, Baltimore JL 20 Penn St. JL Baltimore, MD 21201 JL jalem...@outerbanks.umaryland.edu | (410) 706-7441 JL http://mackerell.umaryland.edu/~jalemkul JL == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] GROMACS and CNT/graphene (probably for the 1000th time)
Great, thanks. One other question, just sort of jumping way ahead. Let's say I had a denaturated protein chain I wanted to drag in a simulation of translocation. Is there anything in Gromacs for that, aside from fake charges and external electric fields, or messing with the code itself? JL Yep. JL -Justin -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
[gmx-users] The largest charge group contains [GIANT CRAPTON] of atoms
I have a custom-FF (OPLSAA based) system with ~8500 atoms, which is a rectangular graphene sheet with a hole and six pieces of small organics covalently bound at the hole edge. About 95% of all atoms are electrically neutral, with the exception of the functional groups mentioned above and a total of 24 graphene edge carbons (-0.06e), each passivated by a hydrogen (+0.06e). All charges have been carefully set (actually, from our own DFT calculations) to make the system electrically neutral. The charge values only have two decimals, set in ffnonbonded. Two issues: 1. x2top yields a total charge of something e-6. I initially ignored it, but now I think it may be indicative of a actual problem, because: 2. grompp produces a fatal error (see subject) with the number of atoms in the group ~half the size of the system. Because I am not using anything out of the box (coordinates, FF), I'm really perplexed here. I checked the manual, and it makes sense that a charge group is a way to lump atoms into locally neutral regions. And that's sort of exactly what I have, at least from the standpoint of geometry. Definitely not worse than a peptide chain common for GMX. Any ideas? Thanks, Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
[gmx-users] GMX-compatible DNA coordinates
I'm fully aware that this is a forum for discussing Gromacs, but there's a bit of a frustrating situation that involves using it. I need a linear ssDNA chain, sequence of my choosing, anything that will work with, say, AMBER99SB-ILDN. Not trying to find any funny forcefields for it, everything as is. You suggested AMBER in the past, so AMBER it is. ChemOffice produced a PDB with incompatible atom types. make-na server (http://structure.usc.edu/make-na/server.html) also produced something pdb2gmx hates. Is there absolutely any way to get a PDB that will produce a usable topology out of the box with an established potential implemented in Gromacs? Anything that will not involve manual modification of PDBs or writing text-processing scripts will make me very happy. I will really appreciate any suggestions. Thank you, Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] GMX-compatible DNA coordinates
Mauricio, That is very useful, thanks a lot. I'll try it in a bit. Alex MEN Alex, MEN You can use X3DNA to produce any nucleic acid geometry you’d like and MEN gromacs pdb2gmx will understand it. MEN For example, to create a single-strand of 16 DNA bases you only need a file MEN with the corresponding rigid-body base-step parameters such as the MEN ssdna.par file shown below: MEN 16 # bases MEN0 # ***local step parameters*** MEN # Shift Slide Rise Tilt Roll Twist MEN T 0.000 0.000 0.000 0.000 0.000 0.000 MEN A 0.280-0.664 3.201 2.499-2.89532.486 MEN C -0.314-0.644 2.640 8.643 2.63128.956 MEN C 0.424-0.849 3.240 1.756 1.69033.400 MEN A 1.519-0.396 3.077 4.184 6.15239.140 MEN A 0.412-0.670 2.797 6.57711.30926.608 MEN C -0.183-2.185 2.859 6.133 5.12617.125 MEN A 0.423-1.477 6.195 -18.545 -26.69136.074 MEN C 1.216-0.879 3.063 5.417 8.46635.858 MEN A -9.111-4.794-5.43374.82943.216 -56.814 MEN C 3.440 0.344 3.176 3.03612.62142.046 MEN A -9.111-4.794-5.43374.82943.216 -56.814 MEN C 3.440 0.344 3.176 3.03612.62142.046 MEN A 1.478 0.681 3.05613.088 0.20844.226 MEN A -0.358-0.777 3.494-2.803-4.53433.162 MEN A -0.264-0.272 2.833 9.637 1.45730.508 MEN With such a file at hand all you have to do is issue the following two MEN commands to 3DNA. MEN x3dna_utils cp_std -d BDNA MEN Which copies a standard BDNA backbone scaffold in the folder you’re at, and MEN then: MEN rebuild -atomic ssdna.par ssdna.pdb MEN This will generate the file ssdna.pdb which you can find at: MEN http://wikisend.com/download/614970/ssdna.pdb MEN It works with: MEN pdb2gmx -f ssdna.pdb -o start.pdb -p topol.top -i posre.itp -ff MEN charmm27 -water tip3p MEN In principle the online version of 3DNA, w3dna, would also work for MEN generating nucleic acid models of any desired sequence. But sadly the X3DNA MEN version working in the backend of the webserver is awfully outdated and MEN uses the old convention of simply A, G, T, C, instead of DA, DG, DT, DC, to MEN name the nucleotide residues. MEN Hope this helps, MEN Mauricio MEN +++ MEN Mauricio Esguerra Neira, Ph.D. MEN Post-Doctoral Fellow MEN Hugo Gutiérrez de Terán P.I. MEN Department of Cell and Molecular Biology MEN Uppsala University MEN Phone Number: (+46)073-678-5334 MEN mesguerra.org MEN mauricio.esgue...@icm.uu.se MEN +++ -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] GMX-compatible DNA coordinates
Hi Tsjerk, You're right. It's just that the last time I used a sed or awk script of my own composition was in college, which is ~13 years ago. :)) So, this is still an undertaking for someone like me. But yes, what you said is well taken. Lack of knowledge/desire to rename atoms in text files shouldn't hinder research. I do think that having a person such as yourself of Justin on my team to implement what I have in mind would speed things up tremendously, but noone seems to be interested, hint-hint. ;) Thanks, Alex Hi Alex, Writing a sed oneliner is not the same as writing a renaming script. I commonly use them to process PDB files in some way and it's surprising how much you can do without scripting. Cheers, Tsjerk On Apr 3, 2015 7:13 PM, "Alex" nedoma...@gmail.com wrote: Gentlemen, Building a script to rename the atom is obviously an option. I was just wondering if I missed something, e.g. a VMD plugin that could do it quickly. As far as google searching, I have done that, and I have found some things (like the make-NA server, or this:http://smog-server.org/cgi-bin/GenTopGro.pl).This is still not quite what I was aiming for. Hence, my initial question. I didn't mean to be snippy, but I did try searching, believe me. On the other hand, I assume that you are much better versed in the problem at hand, so then maybe you'd be able to appreciate info I simply ignored. Obviously, I welcome anything you can suggest. If this boils down to me writing a renaming script, so be it. :) Thanks, Alex On 03/04/2015 5:53 pm, "Mark Abraham" mark.j.abra...@gmail.com wrote: Hi, I tried that search myself and found some things that looked potentially useful to you, so I suggested that search. Either you might find something useful there, or the dialogue could progress about what you need. Requiring perfect input for pdb2gmx is a fairly stringent expectation. i.e. The use of sed or similar tools is normal, as Tsjerk says also. There are too many options, and nobody has built the Babel fish of MD input! Mark Mark On 03/04/2015 3:40 pm, "Alex" nedoma...@gmail.com wrote: Mark, Not to troll, but why did you assume that I didn't do my due diligence before posting here? Because, you see, you could have left that second part of your reply (which only confirms the legitimacy of my initial question), and it would make it less insulting. Alex Hi, Googling things like "gromacs DNA structure" is a good bet. There are no fully standardised naming schemes or workflows, so you will likely not have something do everything for you. Mark On 03/04/2015 7:03 am, "Alex" nedoma...@gmail.com wrote: I'm fully aware that this is a forum for discussing Gromacs, but there's a bit of a frustrating situation that involves using it. I need a linear ssDNA chain, sequence of my choosing, anything that will work with, say, AMBER99SB-ILDN. Not trying to find any funny forcefields for it, everything as is. You suggested AMBER in the past, so AMBER it is. ChemOffice produced a PDB with incompatible atom types. make-na server (http://structure.usc.edu/make-na/server.html) also produced something pdb2gmx hates. Is there absolutely any way to get a PDB that will produce a usable topology out of the box with an established potential implemented in Gromacs? Anything that will not involve manual modification of PDBs or writing text-processing scripts will make me very happy. I will really appreciate any suggestions. Thank you, Alex -- Gromacs Users mailing list * Please search the archive athttp://www.gromacs.org/Support/Mailing_Lists/GMX-Users_Listbefore posting! * Can't post? Readhttp://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-usersor send a mail togmx-users-requ...@gromacs.org. -- Best regards, Alex mailto:nedoma...@gmail.com -- Gromacs Users mailing list * Please search the archive athttp://www.gromacs.org/Support/Mailing_Lists/GMX-Users_Listbefore posting! * Can't post? Readhttp://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-usersor send a mail togmx-users-requ...@gromacs.org. -- Best regards, Alex mailto:nedoma...@gmail.com -- Gromacs Users mailing list * Please search the archive athttp://www.gromacs.org/Support/Mailing_Lists/GMX-Users_Listbefore posting! * Can't post? Readhttp://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-usersor send a mail togmx-users-requ...@gromacs.org. -- Best regards, Alex mailto:nedoma...@gmail.com -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!
Re: [gmx-users] CNT force field and simulation
Great, glad it's working. I think there was a wrong bond length somewhere in Andrea's tutorial, not sure where exactly (I could be wrong). Look at everything carefully and make sure the bond lengths are at 0.142 nm, because this is what the generator is doing. Have fun! Alex Thank you a lot, I did this and exactly as you are saying I do not have the message. I use the link posted in my first message : (http://www.nanotube.msu.edu/tubeASP/) And then, I converted with Babel. But I removed CONECT statements. And now I am modifying files as you said and trying to run the tutorial. Thank you a lot Alex! -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] CNT force field and simulation
Hi Pierre, A standard CNT/graphene parameterization in GMX doesn't currently exist, but if you're just playing around, please follow the Andrea Minoia's tutorial (http://www.webcitation.org/66u2xJJ3O) exactly. It works, and there should be no errors. Make sure your initial PDB is properly formatted. Also, you do not have to create GRO structures. When you box your initial coordinates, you can set the output to .pdb, which x2top and (later on) grompp will accept. There are two issues with that tutorial: 1. The bond, angle, and dihedral constants are wrong, copied from aromatics. I am using constants derived from the optimized Brenner potential. For bond type 1, the constant is 420420.0 kJ/mol/nm^2. The angle constant for type 1 is 659.346 kJ/mol/rad2 (776.923 kJ/mol for type 2). Note that you can't really obtain one constant from the other, angular functions here for small angles are a matter of using Taylor expansion for the original potential from solid-state physics. Finally, for dihedrals (type 3), the constant is 17.30770 kJ/mol/rad2, so modify Andrea's statement to: C C C C 3 17.30770 0.0 -17.30770 0.0 0.0 0.0 The types I mentioned are described in Table 5.5 of the GMX manual. The reference for the parameters above is my own work: http://iopscience.iop.org/0957-4484/25/48/485701 These are far from perfect (mainly because GMX is a molecular mechanics package, so bonds and angles are, well, the simplest possible case in mechanics, i.e. springs), but with these values they're mathematically guaranteed to reproduce the phonon spectrum of graphene/CNTs at 0K. It may not matter for what you're doing, but this is at least some level of rigor, because a stable lattice isn't actually enough. 2. Make sure you set 'constraints = none' in your mdp. Constraint algorithms generally beloved by the biophysical community will turn your nanotubes into solid bricks. Good luck! Alex PB Hi everyone, PB I am new in the Gromacs users community and in the MD field too. However I PB red a lot and I want to realize simulation. I know that it is not trivial PB and that is why I need users help / point of view. PB *Project : *Simulate interactions between a single-walled carbon nanotube PB (SWCNT) and a protein. PB (I red tutorials and mailing list archive about this kind of simulation, PB and that is the reason why I come here to find help). *Tutorial* : Andrea Minoia (*http://www.gromacs.org/Documentation/How-tos/Carbon_Nanotube?highlight=carbon+nanotube PB http://www.gromacs.org/Documentation/How-tos/Carbon_Nanotube?highlight=carbon+nanotube* PB ) PB *Workflow PB I red manual and utilisation of Gromacs, and that is what I did in PB accordance with Andrea's tutorial : PB *#1* PB *-- i)* *PDB file of molecules*. For the SWCNT for example. I generated a PB structure from *Nanotube Generator* PB (*http://www.nanotube.msu.edu/tubeASP/ PB http://www.nanotube.msu.edu/tubeASP/*). From the .xzy file generated, PB with Open-Babel (*http://openbabel.org/wiki/Main_Page PB http://openbabel.org/wiki/Main_Page*). Then, I generated a .pdb file of PB my SWCNT. COMPLETE. PB *-- ii) **Convert in .gro format.* With the command : PB PB editconf -f cnt.pdb -o cnt.gro -box 100 100 100 -angles 90 90 90 PB PB I followed advices and made a larger box than nanotube. PB Warning message : PB PB WARNING: all CONECT records are ignored PB Read 120 atoms PB No velocities found PB system size : 0.783 0.783 1.136 (nm) PB center : 0.000 -0.000 0.568 (nm) PB box vectors : 0.000 0.000 0.000 (nm) PB box angles : 0.00 0.00 0.00 (degrees) PB box volume : 0.00 (nm^3) PB shift : 50.000 50.000 49.432 (nm) PB new center : 50.000 50.000 50.000 (nm) PB new box vectors :100.000100.000100.000 (nm) PB new box angles : 90.00 90.00 90.00 (degrees) PB new box volume :100.00 (nm^3) PB Back Off! I just backed up new_cnt.gro to ./#new_cnt.gro.2# PB gcq#394: Restraint! What possible restraint? (Joseph Conrad) PB So, I think that, I have my atoms but not the connectivity information. PB What should I do? PB Because I think that, I have a .gro file. But the x2top command send an PB error message : Segmentation Fault... And this error is in relation with PB this warning, right? PB This is the first problem. PB *#2* PB Then, the force field for SWNT. PB I red that, I need to edit a force field for the SWCNT to describe PB the behavior of each atoms of this molecule. So, I found that parameters PB what I have to change (Epsilon and sigma of VdW forces, dihedral angles PB etc...). The best way to do this, is to copy oplsaa.ff in the current PB repertory and then modify this new force field. And then, use x2top with PB *-ff* option to use the new force field repertory edited few minutes ago. PB This is the right way? (I can not try it because the first step with PB editconf does not work
Re: [gmx-users] CNT force field and simulation
Hi Pierre, Two things: 1. rtp entries are pretty useless for crystal structures. Biochemical simulations use molecules with a precise atomic composition, so an rtp template makes sense. Not for crystals. Your rtp entry contains useless information and the only thing in your coordinates that can fit that template is one atom, which is already generous. So, you get one atom in your GRO. ;) 2. Andrea isn't using pdb2gmx, which relies on templates. I honestly have no idea why he created that rtp entry, so maybe other users will clarify that. Andrea uses x2top, which will look up data in ffbonded, ffnonbonded, and in atomname to type translator. This is the right way to treat crystals in GMX. Moreover, you will have to do that every time you have a new set of coordinates. Use what he uses and you won't get any new GRO files. All you will get is a topology to go with your original PDB -- and that is all you need to run a simulation. Alex On Mon, Apr 13, 2015 at 5:49 PM, Pierre Bertin bertin...@gmail.com wrote: Hi Alex, All right, I red all this weekend and I know how to modify a force field to add residues (I hope... ahah). I modified OPLSAA force field because I want to add a protein to my CNT. So, I have edited these files with Andrea Minoia datas (and yours) : *aminoacids.rtp* : [ CNT ] 39 [ atoms ] 40 SWTSWT0 2 41 SWTSWT0 3 42 [ bonds ] 43SWT SWT SWT 44SWT SWT I added a new residue named CNT. 2 or 3 bonds, I do not include hydrogen to start. So a carbon can have 2 (extremities) or 3 bonds (internal). *atomname2type.n2t :* 24 CSWT 0 12.011 2SWT 0.142 SWT 0.142 25 CSWT 0 12.011 3SWT 0.142 SWT 0.142 SWT 0.142 Here, conversion of C atoms from CNT.pdb file to SWT new atom type. *atomtypes.atp :* 823 SWT12.01100 ;CNT Here, define the new atom type : SWT with the carbon mass. *ffbonded.itp : * [ bonds ] SWT SWT 10.1420420420.0 ; CNT SWT SWT bond is a bond between 2 carbons in CNT. I have used your data Alex. [ angletype ] SWTSWTSWT 1 120.000659.346 ; CNT Angle type between the 3 atoms of carbon with datas from you email and the same in dihedral with : [ dihedraltypes ] SWTSWTSWTSWT 3 17.30770 0.0 -17.30770 0.0 0.0 0.0 ; CNT I have tried with SWT, with C and in n2t with a reference to oplsaa carbon from benzene. This works, I can build the .gro file with : gmx pdb2gmx -f cnt_gro.pdb -o cnt_gro.gro -p cnt_gro.top And then I select OPLSAA force field and TIP3P for water. Problem : My cnt_gro.gro has only ONE atom of carbon. And I do not have error message. What did I do wrong? Thanks you a lot for your consideration and time. I attach the pdb file. *Pierre BERTIN* Bioinformatics and Biostatistics student (MSc), University Paris-Sud XI, ORSAY. 2015-04-11 14:31 GMT-06:00 Alex nedoma...@gmail.com: Great, glad it's working. I think there was a wrong bond length somewhere in Andrea's tutorial, not sure where exactly (I could be wrong). Look at everything carefully and make sure the bond lengths are at 0.142 nm, because this is what the generator is doing. Have fun! Alex Thank you a lot, I did this and exactly as you are saying I do not have the message. I use the link posted in my first message : ( http://www.nanotube.msu.edu/tubeASP/) And then, I converted with Babel. But I removed CONECT statements. And now I am modifying files as you said and trying to run the tutorial. Thank you a lot Alex! -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] CNT force field and simulation
Also, I would really suggest that you read the Gromacs manual on the subject of topology. This is extremely useful, because there is a philosophical clash between biomolecular simulations and solid-state simulations in terms of the concept of connectivity. You are effectively using a tool that is broadly inappropriate for solid-state systems. So, understanding the concept of topology in GMX-like packages is extremely useful. Gromacs has a manual of outstanding quality, and it should become your Holy Bible. :) Alex On Mon, Apr 13, 2015 at 6:31 PM, Alex nedoma...@gmail.com wrote: Hi Pierre, Two things: 1. rtp entries are pretty useless for crystal structures. Biochemical simulations use molecules with a precise atomic composition, so an rtp template makes sense. Not for crystals. Your rtp entry contains useless information and the only thing in your coordinates that can fit that template is one atom, which is already generous. So, you get one atom in your GRO. ;) 2. Andrea isn't using pdb2gmx, which relies on templates. I honestly have no idea why he created that rtp entry, so maybe other users will clarify that. Andrea uses x2top, which will look up data in ffbonded, ffnonbonded, and in atomname to type translator. This is the right way to treat crystals in GMX. Moreover, you will have to do that every time you have a new set of coordinates. Use what he uses and you won't get any new GRO files. All you will get is a topology to go with your original PDB -- and that is all you need to run a simulation. Alex On Mon, Apr 13, 2015 at 5:49 PM, Pierre Bertin bertin...@gmail.com wrote: Hi Alex, All right, I red all this weekend and I know how to modify a force field to add residues (I hope... ahah). I modified OPLSAA force field because I want to add a protein to my CNT. So, I have edited these files with Andrea Minoia datas (and yours) : *aminoacids.rtp* : [ CNT ] 39 [ atoms ] 40 SWTSWT0 2 41 SWTSWT0 3 42 [ bonds ] 43SWT SWT SWT 44SWT SWT I added a new residue named CNT. 2 or 3 bonds, I do not include hydrogen to start. So a carbon can have 2 (extremities) or 3 bonds (internal). *atomname2type.n2t :* 24 CSWT 0 12.011 2SWT 0.142 SWT 0.142 25 CSWT 0 12.011 3SWT 0.142 SWT 0.142 SWT 0.142 Here, conversion of C atoms from CNT.pdb file to SWT new atom type. *atomtypes.atp :* 823 SWT12.01100 ;CNT Here, define the new atom type : SWT with the carbon mass. *ffbonded.itp : * [ bonds ] SWT SWT 10.1420420420.0 ; CNT SWT SWT bond is a bond between 2 carbons in CNT. I have used your data Alex. [ angletype ] SWTSWTSWT 1 120.000659.346 ; CNT Angle type between the 3 atoms of carbon with datas from you email and the same in dihedral with : [ dihedraltypes ] SWTSWTSWTSWT 3 17.30770 0.0 -17.30770 0.0 0.0 0.0 ; CNT I have tried with SWT, with C and in n2t with a reference to oplsaa carbon from benzene. This works, I can build the .gro file with : gmx pdb2gmx -f cnt_gro.pdb -o cnt_gro.gro -p cnt_gro.top And then I select OPLSAA force field and TIP3P for water. Problem : My cnt_gro.gro has only ONE atom of carbon. And I do not have error message. What did I do wrong? Thanks you a lot for your consideration and time. I attach the pdb file. *Pierre BERTIN* Bioinformatics and Biostatistics student (MSc), University Paris-Sud XI, ORSAY. 2015-04-11 14:31 GMT-06:00 Alex nedoma...@gmail.com: Great, glad it's working. I think there was a wrong bond length somewhere in Andrea's tutorial, not sure where exactly (I could be wrong). Look at everything carefully and make sure the bond lengths are at 0.142 nm, because this is what the generator is doing. Have fun! Alex Thank you a lot, I did this and exactly as you are saying I do not have the message. I use the link posted in my first message : ( http://www.nanotube.msu.edu/tubeASP/) And then, I converted with Babel. But I removed CONECT statements. And now I am modifying files as you said and trying to run the tutorial. Thank you a lot Alex! -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Maximum Force constant for Position Restraints
No, it does not depend on the system content, aside from the mass of the particle. For a simulation requiring numerical integration in time, there is a limit, and I just estimated it above. For the real world, the limit is that there's no such thing as a harmonic position restraint. :) If you want to be really hardcore about it, and still assume there _is_ such a thing as harmonic restraints, then the timestep of the real world is your Planck's time. :) Alex On Tue, Apr 21, 2015 at 1:36 PM, Marcelo Depólo marcelodep...@gmail.com wrote: Thanks Alex. But then, theoretically, is there no limit? All depends on your system's content, right? 2015-04-21 16:16 GMT-03:00 Alex nedoma...@gmail.com: Correction: tau/pi on the left for the highest value and 5*tau/pi for the 10 x period suggestion above. On Tue, Apr 21, 2015 at 1:13 PM, Alex nedoma...@gmail.com wrote: I think this can be estimated from a general physical argument. The absolute max in my opinion should come from 4*pi*tau = sqrt(m/k), where m is the mass of the lightest restrained particle in the system, k is the constant you seek, and tau is the timestep. The coefficient is four because of the Nyquist theorem. I believe GMX tests for high vibrational frequencies according to 20*pi*tau on the left (ten times the period). Hope this helps. Alex On Tue, Apr 21, 2015 at 1:02 PM, Marcelo Depólo marcelodep...@gmail.com wrote: Hi guys! Is there a maximum value of Force Constant for Position restraints? Reading the Manual I could only find the default and the equation to describe it. Best, -- Marcelo Depólo Polêto Group of Structural Bioinformatics - Center of Biotechnology Student of MSc Cell and Molecular Biology - UFRGS (Brazil) B.Sc. Biochemistry - University of Viçosa (Brazil) -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Marcelo Depólo Polêto Group of Structural Bioinformatics - Center of Biotechnology Student of MSc Cell and Molecular Biology - UFRGS (Brazil) B.Sc. Biochemistry - University of Viçosa (Brazil) -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Maximum Force constant for Position Restraints
I think this can be estimated from a general physical argument. The absolute max in my opinion should come from 4*pi*tau = sqrt(m/k), where m is the mass of the lightest restrained particle in the system, k is the constant you seek, and tau is the timestep. The coefficient is four because of the Nyquist theorem. I believe GMX tests for high vibrational frequencies according to 20*pi*tau on the left (ten times the period). Hope this helps. Alex On Tue, Apr 21, 2015 at 1:02 PM, Marcelo Depólo marcelodep...@gmail.com wrote: Hi guys! Is there a maximum value of Force Constant for Position restraints? Reading the Manual I could only find the default and the equation to describe it. Best, -- Marcelo Depólo Polêto Group of Structural Bioinformatics - Center of Biotechnology Student of MSc Cell and Molecular Biology - UFRGS (Brazil) B.Sc. Biochemistry - University of Viçosa (Brazil) -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Maximum Force constant for Position Restraints
Correction: tau/pi on the left for the highest value and 5*tau/pi for the 10 x period suggestion above. On Tue, Apr 21, 2015 at 1:13 PM, Alex nedoma...@gmail.com wrote: I think this can be estimated from a general physical argument. The absolute max in my opinion should come from 4*pi*tau = sqrt(m/k), where m is the mass of the lightest restrained particle in the system, k is the constant you seek, and tau is the timestep. The coefficient is four because of the Nyquist theorem. I believe GMX tests for high vibrational frequencies according to 20*pi*tau on the left (ten times the period). Hope this helps. Alex On Tue, Apr 21, 2015 at 1:02 PM, Marcelo Depólo marcelodep...@gmail.com wrote: Hi guys! Is there a maximum value of Force Constant for Position restraints? Reading the Manual I could only find the default and the equation to describe it. Best, -- Marcelo Depólo Polêto Group of Structural Bioinformatics - Center of Biotechnology Student of MSc Cell and Molecular Biology - UFRGS (Brazil) B.Sc. Biochemistry - University of Viçosa (Brazil) -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] help on Graphene Nano Sheets
I think we've covered this when I was having the same issue. If you're trying to simulate a multi-molecule system, then just copy the entire forcefield .ff folder to your local directory and modify the following files: ffbonded.itp ffnonbonded.itp atomnames2types.n2t No need to create rtp entries for graphene/nanotubes, so you can ignore that part of Andrea's tutorial. Instead of just C in his tutorial, come up with a unique label to avoid conflicts, which should then also be used in your PDB. Then what I would do is create a topology entry for just graphene and convert it to a stand-alone itp by stripping off system definitions. This is basically opening the top file, removing those definitions and resaving as *.itp. After that, you can just use pdb2gmx on the remaining parts of the system and complete the system topology by hand. I know, this sounds like a lot of manual tweaking, but this is the paradigm. Good luck! Alex MC I run a graphene nanolayer, usin the below description (from previous mail. MC Now i got an other error. It look like there is a conflict in a force MC field system, someone can suggest me how to modify it? MC the force field file came from Minoia advices, reported in the web side! MC / MC //Program grompp, VERSION 4.6.7// MC //Source code file: MC /home/marcello/DATI/software/gromacs-4.6.7/src/kernel/topio.c, line: 752// MC // MC //Fatal error:// MC //Syntax error - File forcefield.itp, line 31// MC //Last line read:// MC //'1 3 yes 0.5 0.5'// MC //Found a second defaults directive./* MC * -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] pdb2gmx and protonated states of the residues
Aside from interacting with solvent-dissolved protons (if you add them), the protonation states are permanent in GMX. All forcefields used in these simulators are non-reactive (beyond vdw and electrostatics). Unless you run your simulation in stages, where you manually change protonation states, nothing chanes. Alex B Dear All, B B By H++ and reduce we can have the input PDB for pdb2gmx have the B initial corrected protonated states for some residues, for example, B HIS, ARG, GLU and terminal residues. However biologically speaking with the change of the protein B conformation, the protonated states of the above residues can be changed. B B Will you please introduce to me how GROMACS deal with the possible B protonation state change during the MD process? B B Brett -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] pdb2gmx and periodic molecule
JL The force field parameters for a phosphodiester are the same in a linear, finite JL DNA as they will be in your infinite chain. I didn't mean that anything would need to be changed in the description of residues. The FF handles a finite chain perfectly. JL The problem is in telling pdb2gmx JL don't link i,i+1 here, instead link i,i+(some number of intervening nucleotides). This is exactly what I meant, the linking. Is there a way to achieve this without messing with the code? I just haven't looked at the code, so if you think I'm being a cheeky little bastard, I'll accept that. :) Thanks, Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] pdb2gmx and periodic molecule
Well, none of us in this project have the scripting skills necessary for this, that's sort of the problem... Thanks, Alex Hi Alex, I think the best way is to extend the chain, such that you get an overlap between both ends. So for a stretch of DNA ACGT You would generate TACGTA Then you strip off the terminal atoms and rewire the links over the boundary. It requires renumbering the topology and at first looks a bit cumbersome, but this ensures all the links are correct and the modifications of the termini have no effect, and it's not very hard to script. And, of course, scripting it is a one time effort to avoid doing it manually later. Cheers, Tsjerk On Apr 26, 2015 5:55 AM, "Alex" nedoma...@gmail.com wrote: JL The force field parameters for a phosphodiester are the same in a linear, finite JL DNA as they will be in your infinite chain. I didn't mean that anything would need to be changed in the description of residues. The FF handles a finite chain perfectly. JL The problem is in telling pdb2gmx JL "don't link i,i+1 here, instead link i,i+(some number of intervening nucleotides)." This is exactly what I meant, the linking. Is there a way to achieve this without messing with the code? I just haven't looked at the code, so if you think I'm being a cheeky little bastard, I'll accept that. :) Thanks, Alex -- Gromacs Users mailing list * Please search the archive athttp://www.gromacs.org/Support/Mailing_Lists/GMX-Users_Listbefore posting! * Can't post? Readhttp://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-usersor send a mail togmx-users-requ...@gromacs.org. -- Best regards, Alex mailto:nedoma...@gmail.com -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] pdb2gmx and periodic molecule
Justin, I am not entirely sure I understand your sequence of steps. Why do you want to first run it normally, but then also without adding the hydrogens? The hydrogens (a total of two) are only added at the termini by default, and in the final product there's no need for them anyway. Here's what I'm able to understand: run as usual, delete all bonded data relevant to termini, add bonded descriptions for the residues at the boundary, modify pdb to remove termini (actually, I can generate the periodic ones automatically). Am I missing something that should make my life easier? Also, is there a way to automate this by modifying the forcefield? See, right now (on the periodic structure without termini) pdb2gmx sees the first residue and looks it up as one with a terminus, which is really the gist of the error it throws, so I doubt it even gets to the point of looking at the boundary. Is it that the utility behavior cannot be changed at the ff level, and thus our only option is manual modification of the topology? Thank you, Alex JL pdb2gmx isn't designed to intuitively handle such cases. The best way to go JL about it is: JL 1. Run pdb2gmx normally to add H that you might need and to apply patching such JL that you have a phosphate on one end and an alcohol on the other (presumably you JL already have this last part done and have suitable .tdb entries). JL 2. Delete unnecessary atoms from the termini. JL 3. Run pdb2gmx again with -ter (choose None) and -missing so that it doesn't JL complain at you about dangling termini. JL 4. Manually edit the topology to introduce the needed bond, angles, and dihedrals. JL -Justin -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Maximum Force constant for Position Restraints
David, that's exactly right. For every timestep value, one can derive the appropriate upper limit for the restraint constant in a somewhat physically sound manner. I am just not entirely sure what was the purpose of the initial question, because for infinite restraint constant, I'd just freeze the particle. Actually, come to think of it, my Planck time estimate for Mother Nature's timestep in vibrating bonds is probably incorrect. Effective bond length over the speed of light is more like it. :) Alex DvdS But it boils down to that if you make the force constant very large you DvdS need to make the time step short too which may not be what you want. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] help on Graphene Nano Sheets
Hello, The forcefields for CNT and graphene are identical (for the purpose of this discussion). Two questions: 1. Where did you get the initial graphene structure? 2. Are there any CONECT statements in it? I assume you used Andrea Minoia's guide and the contents of your cnt2_oplsaa.ff directory correspond to what he's got. Alex MC Hi, MC i am trying to model a Graphene nano sheet GNS on Gromacs- MC I follow the advices about the nanotubes presented in the help site. MC _http://www.gromacs.org/Documentation/How-tos/Carbon_Nanotube_ MC I run this command to greate the topology: MC _g_x2top -f graphene.gro -o cnt.top -ff cnt2_oplsaa_ MC I got this error message: MC GROMACS: gmx x2top, VERSION 5.0.4 MC Executable: /usr/local/gromacs/bin/gmx MC Library dir: /usr/local/gromacs/share/gromacs/top MC Command line: MCg_x2top -f graphene.gro -o cnt.top -ff cnt2_oplsaa MC Opening force field file MC /usr/local/gromacs/share/gromacs/top/cnt2_oplsaa.ff/atomnames2types.n2t MC Opening force field file MC /usr/local/gromacs/share/gromacs/top/cnt2_oplsaa.ff/name_topology.n2t MC There are 12 name to type translations in file cnt2_oplsaa.ff MC Generating bonds from distances... MC atom 112 MC There are 1 different atom types in your sample MC Generating angles and dihedrals from bonds... MC segmentazion fault (core dump created) MC Someone can help me on this? MC The graphene force field for nanolayer is the same for nanotube? There MC is something that i have to change? MC bests MC -- MC Marcello Cammarata, Ph.D. MC 3208790796 -- Best regards, Alexmailto:nedoma...@gmail.com -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
[gmx-users] pdb2gmx and periodic molecule
Ahoy, What I have here is a 100% precisely set up periodic ssdna chain of six residues without termini. The goal is to get the 'polymerization' across the box. So, I boxed the chain and set up the periodicity pretty much perfectly. To test, translation in the direction of periodicity by the periodicity constant results in bonds perfectly recognized by things like VMD or pymol (in the concatenated structure). When running pdb2gmx on the boxed structure, the periodicity seems to be ignored and then expectedly I get a fatal error with a custom FF that works fine on finite chains. x2top sees pbc, but we're not using it here. So, no topology in sight. Any ideas? Thanks, Alex p.s. I can post the test structure if there's any doubt about periodicity of the structure itself. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] help on Graphene Nano Sheets
Andrea Minoia's tutorial describes the setup of angles, which in the case of graphene, is extremely simple from the geometry standpoint. There is no need for chancing, just add the proper entries under [dihedraltypes] in ffbonded.itp, x2top will take care of the rest. It's also a pretty good idea to actually read the manual on the definition of angles and dihedrals. Energy minimization destroys the sheet, because there's a mess in your setup. Alex MC Hi people, MC I was able to run the graphene layer sheet, by follow your advices. MC I had a problem with the dihetral angle, MC i solved it by chancing, im the top file the funct parameter from 1 to MC 3, as reported in bold below, MC . MC [ dihedrals ] MC ; aiajakal functc0c1 c2 MC c3c4c5 MC 6 1 2 3 *3 * MC 2 1 6 5 *3 * MC 2 18887 *3* MC 1 2 3 4 *3 * MC ... MC and also, by changing in rtp files all_dihedrals from 1 to 0, MC .. MC [ bondedtypes ] MC ; bonds angles dihedrals impropers all_dihedrals nrexcl HH14 RemoveDih MC 1 1 3 1 *0* 3 1 0 MC .. MC But, when i run the mdrun file, after the minimizzation, the model MC result destroyer, all the atom of graphene are spreads arround monitor, MC and they are not yet connected as in the original geometry. MC I would like to have some advices. What is wrong? how i have to set the MC dihedral angles? MC Thanks. MC Il 23/04/2015 20:04, abhijit Kayal ha scritto: Hi Mercelo Copy the oplsaa.ff directory to your working directory. Then in ffnonbonded.itp file add the following lines.. opls_995 C 612.01100 0.000 A3.4e-01 3.61200e-01 opls_996 C 612.01100 0.000 A3.4e-01 3.61200e-01 opls_997 C 612.01100 0.000 A3.4e-01 3.61200e-01 Then in atomname2type.n2t file add the following lines. Copls_9950 12.011 2C 0.142 C 0.142 Copls_9960 12.011 3C 0.142 C 0.142 C 0.142 Copls_9970 12.011 1C 0.142 Then use g_x2top. This will work. On Thu, Apr 23, 2015 at 10:54 PM, Alex nedoma...@gmail.com wrote: I think we've covered this when I was having the same issue. If you're trying to simulate a multi-molecule system, then just copy the entire forcefield .ff folder to your local directory and modify the following files: ffbonded.itp ffnonbonded.itp atomnames2types.n2t No need to create rtp entries for graphene/nanotubes, so you can ignore that part of Andrea's tutorial. Instead of just C in his tutorial, come up with a unique label to avoid conflicts, which should then also be used in your PDB. Then what I would do is create a topology entry for just graphene and convert it to a stand-alone itp by stripping off system definitions. This is basically opening the top file, removing those definitions and resaving as *.itp. After that, you can just use pdb2gmx on the remaining parts of the system and complete the system topology by hand. I know, this sounds like a lot of manual tweaking, but this is the paradigm. Good luck! Alex MC I run a graphene nanolayer, usin the below description (from previous mail. MC Now i got an other error. It look like there is a conflict in a force MC field system, someone can suggest me how to modify it? MC the force field file came from Minoia advices, reported in the web side! MC / MC //Program grompp, VERSION 4.6.7// MC //Source code file: MC /home/marcello/DATI/software/gromacs-4.6.7/src/kernel/topio.c, line: 752// MC // MC //Fatal error:// MC //Syntax error - File forcefield.itp, line 31// MC //Last line read:// MC //'1 3 yes 0.5 0.5'// MC //Found a second defaults directive./* MC * -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. MC -- MC Marcello Cammarata, Ph.D. MC 3208790796 -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] pdb2gmx and periodic molecule
Yeah, I was trying to avoid that last part. You can tell by now how much I dislike uhm doing work. :) I'll try it, thanks. Alex JL 4. Manually edit the topology to introduce the needed bond, angles, and dihedrals. JL -Justin -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Pull code
Well, no, i don't want two. I want one! :) On Wed, Apr 22, 2015 at 5:17 PM, Ming Tang m21.t...@qut.edu.au wrote: Hi，Alex You can try constraint. In umbrella， both 2 groups move. Sent from my Huawei Mobile Alex nedoma...@gmail.com wrote: Hi all, I have a group (DNA) I'd like to translate relative to the other group (CNT) along the Z-direction so that DNA is the only group that's actually moving. The code I had prior to what I have now worked under GMX 4.5.something. Since the syntax has changed in 5.0.x, I found Justin's old suggestions, and this is what I have now: ;Pull code pull= umbrella pull_geometry = direction pull_coord1_vec = 0.0 0.0 1.0 pull_start = yes pull_coord1_init= 0.0 pull_ngroups= 2 pull_ncoords= 1 pull_coord1_groups = 1 2 pull_group1_name= CNT pull_group2_name= DNA pull_coord1_rate1 = 0.0002 ;nm per ps pull_coord1_k = 5000 ; kJ mol^-1 nm^-2 ;end pull code Does this look right? My simulation isn't nearly done, but I am looking at the Z-displacement in pullx.xvg and it just oscillates around a single value. The simulated time is now about 1.5 ns. Thank you, Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Pull code
Okay, that clarifies it. I wasn't sure, because of your earlier comment, despite the fact that I was using referenced pull. Indeed, in my case, part of the system is essentially solid-state, its COM fixed in space, while other stuff is being moved relative to it. I was also curious how (in more detail than what the manual says) the periodic pull worked. I think I would greatly benefit from the ability to translocate, say, a strand of DNA infinitely across a periodic boundary. As is, that would never work, because the group's COM would be nearly static, eventually resulting in huge forces applied to the pulled group. So, if the periodic option is used, will it be possible to continuously translocate a linear chain with a length comparable to the corresponding box dimension, while the rest of the system is non-translating? Thanks as always. Alex On Wed, Apr 22, 2015 at 6:41 PM, Justin Lemkul jalem...@vt.edu wrote: On 4/22/15 8:37 PM, Alex wrote: That is not an option for me, all groups must exhibit full dynamics, so partial restraint seems like a decent option. It is my understanding that explicitly moving only one group is impossible in the new version. Is that the case? No, you can (and have always been able to) use an absolute restraint relative to the coordinate origin. There are additional complications in this case (noted in the manual), so it's not often used. The current setup that you're using is such that one group is moved with respect to the other. So in effect, you are moving only one group but its position depends on the other. In most cases, the absolute position of the reference group is irrelevant, though not always. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Pull code
That is not an option for me, all groups must exhibit full dynamics, so partial restraint seems like a decent option. It is my understanding that explicitly moving only one group is impossible in the new version. Is that the case? Thanks, Alex On Wed, Apr 22, 2015 at 5:58 PM, Ming Tang m21.t...@qut.edu.au wrote: Hi, Alex I tried this code plus freeze all of the atoms in the group needed to stay in place, and it worked well. -Original Message- From: gromacs.org_gmx-users-boun...@maillist.sys.kth.se [mailto: gromacs.org_gmx-users-boun...@maillist.sys.kth.se] On Behalf Of Alex Sent: Thursday, 23 April 2015 9:47 AM To: Discussion list for GROMACS users Subject: Re: [gmx-users] Pull code The CNT group is actually very strongly restrained, not the entire atomic population, but the edges. So, from inspecting my code, everything looks right? Thanks, Alex Not really. The constraint option keeps a rigid constraint between the two groups. The umbrella keyword specifies a harmonic potential, whose strength is tunable. Otherwise, the two methods are identical. If one group needs to stay in place, it needs position restraints applied to it. If the desired separation is not occurring, either a larger spring force constant or a faster pull rate is needed. -Justin -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] help on Graphene Nano Sheets
Justin, everything works for me with graphene/x2top under the latest version. Am I doing something wrong? :) On Wed, Apr 22, 2015 at 5:26 PM, Justin Lemkul jalem...@vt.edu wrote: On 4/22/15 7:22 PM, marcello cammarata wrote: yes, i used the Minoia tutorial, i use a cnt_oplsaa at the first attempt, and forward i copied and modify the standard oplsaa folder. with both of them i have the same trouble. The seg fault will always occur with version 5.0.4; there is a bug we discovered a few weeks back: http://redmine.gromacs.org/issues/1711 You'll need to use an older version (pre-5.0) for x2top to work properly (or patch the code and upload a fix to solve the bug :). The inizial graphene layer was created using*VMD modelling toolbox*, that was a .pdb file, i convert it in .gro file using *editconf *command. i open the gro file using Avogadro, and it is still a layer. But in the gro file the connections are loose. CONECT entries in PDB files are rarely (never?) used in GROMACS. They are totally irrelevant for the purpose of x2top. -Justin here i report the firt 5 rows of the graphene.gro file. UNNAMED 112 1GRA C1 0.000 0.000 0.000 2GRA C2 -0.123 0.071 0.000 3GRA C3 -0.123 0.213 0.000 4GRA C4 0.000 0.284 0.000 5GRA C5 0.246 0.000 0.000 thanks Il 22/04/2015 19:46, Alex ha scritto: Hello, The forcefields for CNT and graphene are identical (for the purpose of this discussion). Two questions: 1. Where did you get the initial graphene structure? 2. Are there any CONECT statements in it? I assume you used Andrea Minoia's guide and the contents of your cnt2_oplsaa.ff directory correspond to what he's got. Alex MC Hi, MC i am trying to model a Graphene nano sheet GNS on Gromacs- MC I follow the advices about the nanotubes presented in the help site. MC _http://www.gromacs.org/Documentation/How-tos/Carbon_Nanotube_ MC I run this command to greate the topology: MC _g_x2top -f graphene.gro -o cnt.top -ff cnt2_oplsaa_ MC I got this error message: MC GROMACS: gmx x2top, VERSION 5.0.4 MC Executable: /usr/local/gromacs/bin/gmx MC Library dir: /usr/local/gromacs/share/gromacs/top MC Command line: MCg_x2top -f graphene.gro -o cnt.top -ff cnt2_oplsaa MC Opening force field file MC /usr/local/gromacs/share/gromacs/top/cnt2_oplsaa.ff/atomnames2types.n2t MC Opening force field file MC /usr/local/gromacs/share/gromacs/top/cnt2_oplsaa.ff/name_topology.n2t MC There are 12 name to type translations in file cnt2_oplsaa.ff MC Generating bonds from distances... MC atom 112 MC There are 1 different atom types in your sample MC Generating angles and dihedrals from bonds... MC segmentazion fault (core dump created) MC Someone can help me on this? MC The graphene force field for nanolayer is the same for nanotube? There MC is something that i have to change? MC bests MC -- MC Marcello Cammarata, Ph.D. MC 3208790796 -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Pull code
The CNT group is actually very strongly restrained, not the entire atomic population, but the edges. So, from inspecting my code, everything looks right? Thanks, Alex Not really. The constraint option keeps a rigid constraint between the two groups. The umbrella keyword specifies a harmonic potential, whose strength is tunable. Otherwise, the two methods are identical. If one group needs to stay in place, it needs position restraints applied to it. If the desired separation is not occurring, either a larger spring force constant or a faster pull rate is needed. -Justin -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
[gmx-users] Pull code
Hi all, I have a group (DNA) I'd like to translate relative to the other group (CNT) along the Z-direction so that DNA is the only group that's actually moving. The code I had prior to what I have now worked under GMX 4.5.something. Since the syntax has changed in 5.0.x, I found Justin's old suggestions, and this is what I have now: ;Pull code pull= umbrella pull_geometry = direction pull_coord1_vec = 0.0 0.0 1.0 pull_start = yes pull_coord1_init= 0.0 pull_ngroups= 2 pull_ncoords= 1 pull_coord1_groups = 1 2 pull_group1_name= CNT pull_group2_name= DNA pull_coord1_rate1 = 0.0002 ;nm per ps pull_coord1_k = 5000 ; kJ mol^-1 nm^-2 ;end pull code Does this look right? My simulation isn't nearly done, but I am looking at the Z-displacement in pullx.xvg and it just oscillates around a single value. The simulated time is now about 1.5 ns. Thank you, Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] help on Graphene Nano Sheets
You don't need to run x2top on the GRO file, you can apply it directly to the PDB. It seems like the PDB to GRO conversion destroyed the coordinates due to poor formatting of the initial coordinates. This has nothing to do with the forcefields or graphene, you may just have a crappy PDB. Alex On Wed, Apr 22, 2015 at 5:22 PM, marcello cammarata marcello.cammar...@unipa.it wrote: yes, i used the Minoia tutorial, i use a cnt_oplsaa at the first attempt, and forward i copied and modify the standard oplsaa folder. with both of them i have the same trouble. The inizial graphene layer was created using* VMD modelling toolbox*, that was a .pdb file, i convert it in .gro file using *editconf *command. i open the gro file using Avogadro, and it is still a layer. But in the gro file the connections are loose. here i report the firt 5 rows of the graphene.gro file. UNNAMED 112 1GRA C1 0.000 0.000 0.000 2GRA C2 -0.123 0.071 0.000 3GRA C3 -0.123 0.213 0.000 4GRA C4 0.000 0.284 0.000 5GRA C5 0.246 0.000 0.000 thanks Il 22/04/2015 19:46, Alex ha scritto: Hello, The forcefields for CNT and graphene are identical (for the purpose of this discussion). Two questions: 1. Where did you get the initial graphene structure? 2. Are there any CONECT statements in it? I assume you used Andrea Minoia's guide and the contents of your cnt2_oplsaa.ff directory correspond to what he's got. Alex MC Hi, MC i am trying to model a Graphene nano sheet GNS on Gromacs- MC I follow the advices about the nanotubes presented in the help site. MC _http://www.gromacs.org/Documentation/How-tos/Carbon_Nanotube_ MC I run this command to greate the topology: MC _g_x2top -f graphene.gro -o cnt.top -ff cnt2_oplsaa_ MC I got this error message: MC GROMACS: gmx x2top, VERSION 5.0.4 MC Executable: /usr/local/gromacs/bin/gmx MC Library dir: /usr/local/gromacs/share/gromacs/top MC Command line: MCg_x2top -f graphene.gro -o cnt.top -ff cnt2_oplsaa MC Opening force field file MC /usr/local/gromacs/share/gromacs/top/cnt2_oplsaa.ff/atomnames2types.n2t MC Opening force field file MC /usr/local/gromacs/share/gromacs/top/cnt2_oplsaa.ff/name_topology.n2t MC There are 12 name to type translations in file cnt2_oplsaa.ff MC Generating bonds from distances... MC atom 112 MC There are 1 different atom types in your sample MC Generating angles and dihedrals from bonds... MC segmentazion fault (core dump created) MC Someone can help me on this? MC The graphene force field for nanolayer is the same for nanotube? There MC is something that i have to change? MC bests MC -- MC Marcello Cammarata, Ph.D. MC 3208790796 -- Marcello Cammarata, Ph.D.3208790796 -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Problems in the modelling of interaction between peptide and copper
Hi, Your model is not reasonable. First of all, let's forget about simulations for a moment, because GMX is a wildly inappropriate choice. Even if the energy was not zero (which in your case is probably due to some LJ parameters set up improperly, or some really good GMX code giving you a hint that this is a bad idea), the results would describe a protein next to a collection of LJ entities put together in a crystal lattice. Here is why: When exposed to air or water, copper surface would undergo quick oxidation. That aside, let us say we have a magic copper surface, which is not oxidated. This is even more problematic for MD-type methods, because metal surfaces don't really have conventional van der Waals due to presence of Fermi gas in metallic crystals, as opposed to the type of polarization in covalent dielectrics or even semiconductors. The only method somewhat reasonable for metals in MD I am aware of is the Modified Embedded Atom Method (MEAM), which, I believe, is implemented in LAMMPS. At the same time, it won't handle the protein. Alternatively, you can use DFTMD in CP2K and actually place the short region of your protein close to the surface of copper and see what happens. This in fact appears to be a half-decent idea and I'd try it. Depending on the exchange correlation functions and the basis you choose, you could probably do about a 100 atoms for a few picoseconds within reasonable time with parallelization over 40-50 decent cores. One thing to keeo in mind is that the last time a checked (a few months back), CP2K had no electron energy sampling aside from the Gamma point (VASP can do that, but it is expensive). If that doesn't bother you, I would recommend going the DFT route. You could even try to build the energy surface for your protein-metal interaction and then try to use it to parameterize a lower level method such as MD. Hope this helps. Alex MT Dear all, MT I got a problem when modelling the interaction of a peptide and MT copper. I am using the force field of opls-aa. The charge of MT copper atoms are defined as zero and the non-bonded parameters are MT obtained from the CU ions, which can be MT found in opls-aa. MT I applied smd and found that the peptide is approaching the MT copper surface and stopped when it is very close to the surface, MT due to some abnormal interaction. The whole process seems quite MT reasonable, however, when I rerun the MT simulation to obtain the interaction energy, for the vdw MT interaction between peptide and copper is zero, absolute zero, in the whole process. MT Can anyone give me some suggestions? Is my model reasonable? I am MT not so confident with modelling strategy of using the LJ MT parameters for copper bulk material. Also the zero interaction energy is another big issue. MT Cheers, MT Tng -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Problems in the modelling of interaction between peptide and copper
Also, I assumed by copper you meant an actual surface of copper, not a few copper ions floating in water. :) Alex MT Dear all, MT I got a problem when modelling the interaction of a peptide and MT copper. I am using the force field of opls-aa. The charge of MT copper atoms are defined as zero and the non-bonded parameters are MT obtained from the CU ions, which can be MT found in opls-aa. MT I applied smd and found that the peptide is approaching the MT copper surface and stopped when it is very close to the surface, MT due to some abnormal interaction. The whole process seems quite MT reasonable, however, when I rerun the MT simulation to obtain the interaction energy, for the vdw MT interaction between peptide and copper is zero, absolute zero, in the whole process. MT Can anyone give me some suggestions? Is my model reasonable? I am MT not so confident with modelling strategy of using the LJ MT parameters for copper bulk material. Also the zero interaction energy is another big issue. MT Cheers, MT Tng -- Best regards, Alexmailto:nedoma...@gmail.com -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Constraints applied for keeping rigid in ND
Is there a particular reason you're applying LINCS constraints to your diamond structure? Usually, LINCS gives convergence errors/warnings if the system is poorly equilibrated and/or your time step is too large. The fact that your system was soft without constraints actually suggests poor setup, which once again could include the time step. Diamond structure, if properly set up, should be quite rigid, even if your bond/angle terms are somewhat off, because, well, it's the diamond structure. Alex hy Hi all, hy My research focuses on the dynamics of a nanodiamond(ND) in a biological hy environment. I am working with gromacs version 5.0.2 and using Martini CG hy ff. In order to keep the rigidity of ND, I applied constraints as the stiff hy bond in the .itp file. But my system crashed in the NVT run with LINCS hy WARNING. So I tried to use a small ND, an octahedron, in the water box for hy debugging. It did work well. However, after I increased its size(just add 3 hy beads actually), it crashed with same input parameters. hy How it could happen? I am the beginner using gromacs and CG model. I think hy the problem is due to the constraints because the same system with defined hy bond interaction worked well(but the ND in this situation is soft). I am hy just wondering that is there any limitation for applying constraints for hy crystal packing system or anything wrong with my configuration? hy Seeking for help. Thanks a lot. hy Kevin hy ** hy My itp file: hy ;itp file of ND hy [ moleculetype ] hy ; molname nrexcl hy CG-ND 1 hy [ atoms ] hy ; id type resnr residu atom cgnr charge hy 1 P1 1 FO CD1 1 0.00 hy 2 C1 1 FH CD2 2 0.00 hy 3 P1 1 FO CD3 3 0.00 hy 4 C1 1 FH CD4 4 0.00 hy 5 C1 2 FH CD1 5 0.00 hy 6 C1 2 FH CD2 6 0.00 hy 7 C1 2 FH CD3 7 0.00 hy 8 P1 2 FO CD4 8 0.00 hy 9 P1 3 FO CD1 9 0.00 hy 10 P1 3 FO CD2 100.00 hy [ constraints ] hy ; i j funct length force.c. hy #ifdef FLEXIBLE hy [ bonds ] hy #endif hy 12 1 0.317 100 hy 13 1 0.317 100 hy 14 1 0.317 100 hy 23 1 0.317 100 hy 24 1 0.317 100 hy 34 1 0.317 100 hy 35 1 0.317 100 hy 36 1 0.317 100 hy 45 1 0.317 100 hy 46 1 0.317 100 hy 56 1 0.317 100 hy 57 1 0.317 100 hy 58 1 0.317 100 hy 67 1 0.317 100 hy 68 1 0.317 100 hy 710 1 0.317 100 hy 78 1 0.317 100 hy 79 1 0.317 100 hy 810 1 0.317 100 hy 89 1 0.317 100 hy 910 1 0.317 100 hy ** hy My mdp file: hy title = debugging for ND hy ; Run parameters hy integrator = md ; leap-frog integrator hy nsteps = 2500; 20 * 2500 = 500 ps (0.05 ns) hy dt= 0.02 ; 20 fs hy ; Output control hy nstxout = 1 ; save coordinates hy nstvout = 1 ; save velocities hy nstenergy = 10 ; save energies hy nstlog = 10 ; update log file hy nstcomm = 10 hy fcstep = 0 hy nstcalcenergy = 10 hy ; Bond parameters hy unconstrained_start = no hy constraints = none hy constraint_algorithm = Lincs hy lincs_order = 4 hy lincs_warnangle = 30 hy lincs-iter = 1 hy morse= no hy ; Neighborsearching hy cutoff-scheme= verlet hy verlet-buffer-drift = 0.001 hy ns_type = grid ; search neighboring grid cels hy nstlist = 10; hy rlist = 1.4 ; short-range neighborlist cutoff (in nm) hy rcoulomb = 1.2 ; short-range electrostatic cutoff (in nm) hy rvdw = 1.2 ; short-range van der Waals cutoff (in nm) hy vdw_type= cutoff hy rvdw_switch = 0.9 hy vdw-modifier = Potential-shift hy ; Electrostatics hy coulombtype = Reaction_field hy coulomb_modifier= Potential-shift hy rcoulomb_switch = 0.0 hy epsilon_r = 15 hy epsilon_rf = 0 hy ; Temperature coupling is on hy tcoupl = Berendsen hy tc-grps = system ; three coupling groups - more accurate hy tau_t = 1.0; time constant, in ps hy ref_t = 310 ; reference temperature, one for each group, in K hy ; Pressure coupling is on hy pcoupl = no; Pressure coupling on in NPT hy pcoupltype = isotropic; uniform scaling of x-y-z box vectors hy tau_p = 4.0; time constant hy ref_p = 1.0; reference pressure hy compressibility = 1e-5 ; isothermal compressibility, bar^-1 hy ; Periodic boundary conditions hy pbc= xyz ; 3-D PBC hy ; Velocity generation hy gen_vel = no hy gen_temp = 310 hy gen_seed = -1 -- Gromacs Users mailing list * Please search the archive
Re: [gmx-users] Constraints applied for keeping rigid in ND
Also, I just looked at your topology. Where are your angles? Try to implement them and then turn off constraints. Alex A Is there a particular reason you're applying LINCS constraints to your A diamond structure? Usually, LINCS gives convergence A errors/warnings if the system is poorly A equilibrated and/or your time step is too large. The fact that your A system was soft without constraints actually suggests poor setup, A which once again could include the time step. A Diamond structure, if properly set up, should be quite rigid, even if A your bond/angle terms are somewhat off, because, well, it's the A diamond structure. A Alex hy Hi all, hy My research focuses on the dynamics of a nanodiamond(ND) in a biological hy environment. I am working with gromacs version 5.0.2 and using Martini CG hy ff. In order to keep the rigidity of ND, I applied constraints as the stiff hy bond in the .itp file. But my system crashed in the NVT run with LINCS hy WARNING. So I tried to use a small ND, an octahedron, in the water box for hy debugging. It did work well. However, after I increased its size(just add 3 hy beads actually), it crashed with same input parameters. hy How it could happen? I am the beginner using gromacs and CG model. I think hy the problem is due to the constraints because the same system with defined hy bond interaction worked well(but the ND in this situation is soft). I am hy just wondering that is there any limitation for applying constraints for hy crystal packing system or anything wrong with my configuration? hy Seeking for help. Thanks a lot. hy Kevin hy ** hy My itp file: hy ;itp file of ND hy [ moleculetype ] hy ; molname nrexcl hy CG-ND 1 hy [ atoms ] hy ; id type resnr residu atom cgnr charge hy 1 P1 1 FO CD1 1 0.00 hy 2 C1 1 FH CD2 2 0.00 hy 3 P1 1 FO CD3 3 0.00 hy 4 C1 1 FH CD4 4 0.00 hy 5 C1 2 FH CD1 5 0.00 hy 6 C1 2 FH CD2 6 0.00 hy 7 C1 2 FH CD3 7 0.00 hy 8 P1 2 FO CD4 8 0.00 hy 9 P1 3 FO CD1 9 0.00 hy 10 P1 3 FO CD2 100.00 hy [ constraints ] hy ; i j funct length force.c. hy #ifdef FLEXIBLE hy [ bonds ] hy #endif hy 12 1 0.317 100 hy 13 1 0.317 100 hy 14 1 0.317 100 hy 23 1 0.317 100 hy 24 1 0.317 100 hy 34 1 0.317 100 hy 35 1 0.317 100 hy 36 1 0.317 100 hy 45 1 0.317 100 hy 46 1 0.317 100 hy 56 1 0.317 100 hy 57 1 0.317 100 hy 58 1 0.317 100 hy 67 1 0.317 100 hy 68 1 0.317 100 hy 710 1 0.317 100 hy 78 1 0.317 100 hy 79 1 0.317 100 hy 810 1 0.317 100 hy 89 1 0.317 100 hy 910 1 0.317 100 hy ** hy My mdp file: hy title = debugging for ND hy ; Run parameters hy integrator = md ; leap-frog integrator hy nsteps = 2500; 20 * 2500 = 500 ps (0.05 ns) hy dt= 0.02 ; 20 fs hy ; Output control hy nstxout = 1 ; save coordinates hy nstvout = 1 ; save velocities hy nstenergy = 10 ; save energies hy nstlog = 10 ; update log file hy nstcomm = 10 hy fcstep = 0 hy nstcalcenergy = 10 hy ; Bond parameters hy unconstrained_start = no hy constraints = none hy constraint_algorithm = Lincs hy lincs_order = 4 hy lincs_warnangle = 30 hy lincs-iter = 1 hy morse= no hy ; Neighborsearching hy cutoff-scheme= verlet hy verlet-buffer-drift = 0.001 hy ns_type = grid ; search neighboring grid cels hy nstlist = 10; hy rlist = 1.4 ; short-range neighborlist cutoff (in nm) hy rcoulomb = 1.2 ; short-range electrostatic cutoff (in nm) hy rvdw = 1.2 ; short-range van der Waals cutoff (in nm) hy vdw_type= cutoff hy rvdw_switch = 0.9 hy vdw-modifier = Potential-shift hy ; Electrostatics hy coulombtype = Reaction_field hy coulomb_modifier= Potential-shift hy rcoulomb_switch = 0.0 hy epsilon_r = 15 hy epsilon_rf = 0 hy ; Temperature coupling is on hy tcoupl = Berendsen hy tc-grps = system ; three coupling groups - more accurate hy tau_t = 1.0; time constant, in ps hy ref_t = 310 ; reference temperature, one for each group, in K hy ; Pressure coupling is on hy pcoupl = no; Pressure coupling on in NPT hy pcoupltype = isotropic; uniform scaling of x-y-z box vectors hy tau_p = 4.0; time constant hy ref_p = 1.0; reference pressure hy compressibility = 1e-5 ; isothermal compressibility, bar^-1 hy ; Periodic boundary conditions hy pbc= xyz ; 3-D PBC hy ; Velocity generation hy
Re: [gmx-users] Supercritical CO2
Unfortunately, I had a pretty terrible experience trying to simulate fuels, fuel mixtures, and CO2 dissolved in fuels with Gromacs (using OPLS-AA) at high pressures and/or temperatures. It is my understanding that at the moment this is isn't the area where these forcefields are seriously tested. I have no results for other forcefields, but for OPLS-AA it was unpublishable and came out as an internal note at the institution. If curious, see http://nvlpubs.nist.gov/nistpubs/TechnicalNotes/NIST.TN.1805.pdf In fact, if you get any results that correspond to experimental data, I'll really appreciate it if you let me know. :) Alex On Thu, Apr 30, 2015 at 2:04 PM, Jarrett Lee Wise jwi...@uwyo.edu wrote: Hello all- I am trying to run simulations containing supercritical CO2 and oil asphaltenes. I have tested various force fields (TraPPE flex, TraPPe rigid, Charmm, EPM2, Cygan, and Zhang), but all of my results give either extremely low densities (~200 kg/m^3) or extremely high densities (~1600 kg/m^3) when the density should be around 630 kg/m^3. I equilibrated a system of 5000 CO2 molecules with Nose-Hoover and Berendsen for 5 ns, then did a production run using Nose-Hoover and Parrinello-Rahman for another 5 ns using 1 fs time step. I have found that tau_p plays an important factor. A relaxed tau_p gives low density while a vigorous tau_p gives a high density and crashes using Parrinello-Rahman pressure coupling. Has anyone else encountered similar problems or have any advice? Here is a copy of my MDP file. Any advice is greatly appreciated. Thanks Jarrett Wise PhD Student Petroleum Engineering University of Wyoming title = CO2 Density Run integrator = md nsteps = 500 dt = 0.001 nstxout = 0 nstvout = 0 nstfout = 0 nstxtcout = 2 nstenergy = 100 nstlog = 1500 constraint_algorithm = lincs constraints = H-Bonds lincs_iter = 1 lincs_order = 4 cutoff-scheme = Verlet ns_type = grid nstlist = 100 rlist = 1.4 rcoulomb= 1.4 rvdw= 1.4 nstcalcenergy = 10 comm_mode = linear coulombtype = PME pme_order = 4 fourierspacing = 0.16 tcoupl = Nose-Hoover tc-grps = System tau_t = 5.0 ref_t = 320 pcoupl = Parrinello-Rahman pcoupltype = isotropic tau_p = 3.0 ref_p = 520.0 compressibility = 1.45e-4 pbc = xyz continuation= yes gen_vel = no gen_temp = 320 -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Grompp error No default Ryckaert-Bell. types
"Which angles?" -- wait until Mark Abraham gets here. You must have a very special version of aminoacids.rtp Example: https://github.com/gromacs/gromacs/blob/master/share/top/gromos43a2.ff/aminoacids.rtp Thank you But which angles I need to add in .rtp file. are you talking about impropers and dihedrals ??? as In aminoacid.rtp there is no section for angles. Best Anurag Dobhal Juniour Research fellow Supervisor: Dr. Ratnesh D. Jain Department of chemical engineering Insttute of Chemical Technology 400019, +91 8898486877 On Mon, May 4, 2015 at 12:47 PM, Sasha nedoma...@gmail.com wrote: You said you found the rtp entry to be correct, and yet the one you attached has no angles listed, none whatsoever. Alex Thank you for your reply. I am simulating a PLGA (poly lactic co-glycolic acid molecle). I get the topology file and I solvated the molecule. my system is neutral. while running the grompp I got 6 No default Ryckaert-Bell. types. I rechecked my parameters in .rtp file and I found them correct, or may be I am unable to found the problem. I am using ions.mdp file which is given in the tutorial (lysozyme tutorial). I have attatched the .pdb file and the .rtp parameters for them. can you please check the parameters which I have given. so that I can rectify the errors. awaiting for your reply. Best Anurag Dobhal Juniour Research fellow Supervisor: Dr. Ratnesh D. Jain Department of chemical engineering Insttute of Chemical Technology 400019, +91 8898486877 On Mon, May 4, 2015 at 12:11 PM, Alex nedoma...@gmail.com wrote: It may sound surprising, but this is resolved by either introducing correct R-B term parameters in your topology/ff definitions, or by removing the mess from the input coordinates, especially when non-bonded atoms are determined to participate in a dihedral. Humor aside, more information about the system is needed. Alex AD Hello all AD How can I address the "No default Ryckaert-Bell. types error". AD The error is also in between the atoms which are not bonded with each other. AD Thank you AD -- AD *DISCLAIMER:* AD *This communication is intended only for the person or entity to which it AD is addressed and may contain confidential and / or privileged material. Any AD review, retransmission, dissemination or other use is prohibited. If you AD have received this in error, please contact the sender and delete this AD material from your computer. Kindly note that this mail does not constitute AD an offer or solicitation for the purchase or sale of any financial AD instrument or as an official confirmation of any transaction. The AD information is not warranted as to completeness or accuracy and is subject AD to change without notice. Any comments or statements made herein do not AD necessarily reflect those of Nanomedicine Research Group. Before opening AD the email or accessing any attachments, please check and scan for virus.* DISCLAIMER: This communication is intended only for the person or entity to which it is addressed and may contain confidential and / or privileged material. Any review, retransmission, dissemination or other use is prohibited. If you have received this in error, please contact the sender and delete this material from your computer. Kindly note that this mail does not constitute an offer or solicitation for the purchase or sale of any financial instrument or as an official confirmation of any transaction. The information is not warranted as to completeness or accuracy and is subject to change without notice. Any comments or statements made herein do not necessarily reflect those of Nanomedicine Research Group. Before opening the email or accessing any attachments, please check and scan for virus. -- Best regards, Sasha mailto:nedoma...@gmail.com DISCLAIMER: This communication is intended only for the person or entity to which it is addressed and may contain confidential and / or privileged material. Any review, retransmission, dissemination or other use is prohibited. If you have received this in error, please contact the sender and delete this material from your computer. Kindly note that this mail does not constitute an offer or solicitation for the purchase or sale of any financial instrument or as an official confirmation of any transaction. The information is not warranted as to completeness or accuracy and is subject to change without notice. Any comments or statements made herein do not necessarily reflect those of Nanomedicine Research Group. Before opening the email or accessing any attachments, please check and scan for virus. -- Best regards, Alex mailto:nedoma...@gmail.com -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth
Re: [gmx-users] Grompp error No default Ryckaert-Bell. types
The rtp file format and its acceptable entries are not forcefield-specific (oplsaa seems to look up triplets for [ angles ] elsewhere), I just wanted to give an example of that. The oplsaa version of aminoacids.rtp only lists dihedral and improper angles explicitly, and that is exactly what's grompp complaining about in your case. Alex Dear Sir I am using oplsaa force field. In the .rtp file i did not found angles secion for any amino acid. yes there are sections for dihedrals and impropers. sorry for troubling you. Best Anurag Dobhal Juniour Research fellow Supervisor: Dr. Ratnesh D. Jain Department of chemical engineering Insttute of Chemical Technology 400019, +91 8898486877 On Mon, May 4, 2015 at 1:03 PM, Alex nedoma...@gmail.com wrote: "Which angles?" -- wait until Mark Abraham gets here. You must have a very special version of aminoacids.rtp Example:https://github.com/gromacs/gromacs/blob/master/share/top/gromos43a2.ff/aminoacids.rtp Thank you But which angles I need to add in .rtp file. are you talking about impropers and dihedrals ??? as In aminoacid.rtp there is no section for angles. Best Anurag Dobhal Juniour Research fellow Supervisor: Dr. Ratnesh D. Jain Department of chemical engineering Insttute of Chemical Technology 400019, +91 8898486877 On Mon, May 4, 2015 at 12:47 PM, Sasha nedoma...@gmail.com wrote: You said you found the rtp entry to be correct, and yet the one you attached has no angles listed, none whatsoever. Alex Thank you for your reply. I am simulating a PLGA (poly lactic co-glycolic acid molecle). I get the topology file and I solvated the molecule. my system is neutral. while running the grompp I got 6 No default Ryckaert-Bell. types. I rechecked my parameters in .rtp file and I found them correct, or may be I am unable to found the problem. I am using ions.mdp file which is given in the tutorial (lysozyme tutorial). I have attatched the .pdb file and the .rtp parameters for them. can you please check the parameters which I have given. so that I can rectify the errors. awaiting for your reply. Best Anurag Dobhal Juniour Research fellow Supervisor: Dr. Ratnesh D. Jain Department of chemical engineering Insttute of Chemical Technology 400019, +91 8898486877 On Mon, May 4, 2015 at 12:11 PM, Alex nedoma...@gmail.com wrote: It may sound surprising, but this is resolved by either introducing correct R-B term parameters in your topology/ff definitions, or by removing the mess from the input coordinates, especially when non-bonded atoms are determined to participate in a dihedral. Humor aside, more information about the system is needed. Alex AD Hello all AD How can I address the "No default Ryckaert-Bell. types error". AD The error is also in between the atoms which are not bonded with each other. AD Thank you AD -- AD *DISCLAIMER:* AD *This communication is intended only for the person or entity to which it AD is addressed and may contain confidential and / or privileged material. Any AD review, retransmission, dissemination or other use is prohibited. If you AD have received this in error, please contact the sender and delete this AD material from your computer. Kindly note that this mail does not constitute AD an offer or solicitation for the purchase or sale of any financial AD instrument or as an official confirmation of any transaction. The AD information is not warranted as to completeness or accuracy and is subject AD to change without notice. Any comments or statements made herein do not AD necessarily reflect those of Nanomedicine Research Group. Before opening AD the email or accessing any attachments, please check and scan for virus.* DISCLAIMER: This communication is intended only for the person or entity to which it is addressed and may contain confidential and / or privileged material. Any review, retransmission, dissemination or other use is prohibited. If you have received this in error, please contact the sender and delete this material from your computer. Kindly note that this mail does not constitute an offer or solicitation for the purchase or sale of any financial instrument or as an official confirmation of any transaction. The information is not warranted as to completeness or accuracy and is subject to change without notice. Any comments or statements made herein do not necessarily reflect those of Nanomedicine Research Group. Before opening the email or accessing any attachments, please check and scan for virus. -- Best regards, Sasha mailto:nedoma...@gmail.com DISCLAIMER: This communication is intended only for the person or entity to which it is addressed and may contain confidential and / or privileged material. Any review, retransmission, dissemination or other use is prohibited. If you have received this in error, please contact the sender and delete this material from your computer. Kindly note that this mail does not
Re: [gmx-users] Grompp error No default Ryckaert-Bell. types
JL 2. As far as bonded interactions are concerned, in the .rtp only [bonds] are JL required to define connectivity. Remaining bonded interactions are then deduced JL from bond connectivity. If a directive like [angles] or [dihedrals] is present, JL those values are used, instead. If an [angles] directive is missing that just JL tells pdb2gmx create all possible angles from the given bonded connectivity. Well, yes, but the atomtypes were defined in that rtp and no correct triplets or quadruplets were available in ffbonded for those combinations. Then, either define directly in rtp, or stick them into your ffbonded. Am I missing something here? Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Grompp error No default Ryckaert-Bell. types
True, thats why i mentioned Mark.:-) In all honesty, though, for small molecules, i would prefer to ignore the new format and actually add everything in the rtp entry to make ffbonded look like less of a zoo. Also, it was late night of extreme gromacs intercourse for me, which finally paid off. :-) Alex On May 4, 2015 11:31 AM, Justin Lemkul jalem...@vt.edu wrote: On 5/4/15 1:06 PM, Alex wrote: JL 2. As far as bonded interactions are concerned, in the .rtp only [bonds] are JL required to define connectivity. Remaining bonded interactions are then deduced JL from bond connectivity. If a directive like [angles] or [dihedrals] is present, JL those values are used, instead. If an [angles] directive is missing that just JL tells pdb2gmx create all possible angles from the given bonded connectivity. Well, yes, but the atomtypes were defined in that rtp and no correct triplets or quadruplets were available in ffbonded for those combinations. Then, either define directly in rtp, or stick them into your ffbonded. Am I missing something here? I was just clarifying what was a very (to me) confusing exchange. The statements made could easily be taken out of context and misapplied. oplsaa seems to look up triplets for [ angles ] elsewhere is not really right, that's what I was explaining about the fundamental way pdb2gmx processes the .rtp to write angles and dihedrals. The oplsaa version of aminoacids.rtp only lists dihedral and improper angles explicitly, and that is exactly what's grompp complaining about in your case. As worded, this seems like you're saying that the fact that only dihedrals and impropers are listed is the source of the problem. That's not the case, but what you say above is - the parameters for the defined interaction are simply not in the force field and therefore must be added. Of course, if the OP had simply Googled the error, this information would have been readily apparent and would have saved all involved a lot of typing :) -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Supercritical CO2
With opls-aa, incorrect densities and diffusivities were calculated entirely away from the critical point for several liquids. Alex On May 1, 2015 11:40 AM, V.V.Chaban vvcha...@gmail.com wrote: This problem is natural. Even small change in (T,P) in the supercritical range results in a huge change of density and, hence, all other properties. You need either to alter (T,P) or to adjust the force field model (both change will be slight, but will bring large changes of density). If you expect any phase separation in these setups, take care of the Gibbs rule. On Thu, Apr 30, 2015 at 6:38 PM, Alex nedoma...@gmail.com wrote: Only other suggestion I have is that 5 ns production runs are probably not long enough. Granted, we needed diffusion data, so I believe I had 50 to 100 ns long simulations, but even for the barostat to settle, it may be a good idea to have simulations significantly longer than 5 ns. Alex On Thu, Apr 30, 2015 at 3:14 PM, Alex nedoma...@gmail.com wrote: Unfortunately, I had a pretty terrible experience trying to simulate fuels, fuel mixtures, and CO2 dissolved in fuels with Gromacs (using OPLS-AA) at high pressures and/or temperatures. It is my understanding that at the moment this is isn't the area where these forcefields are seriously tested. I have no results for other forcefields, but for OPLS-AA it was unpublishable and came out as an internal note at the institution. If curious, see http://nvlpubs.nist.gov/nistpubs/TechnicalNotes/NIST.TN.1805.pdf In fact, if you get any results that correspond to experimental data, I'll really appreciate it if you let me know. :) Alex On Thu, Apr 30, 2015 at 2:04 PM, Jarrett Lee Wise jwi...@uwyo.edu wrote: Hello all- I am trying to run simulations containing supercritical CO2 and oil asphaltenes. I have tested various force fields (TraPPE flex, TraPPe rigid, Charmm, EPM2, Cygan, and Zhang), but all of my results give either extremely low densities (~200 kg/m^3) or extremely high densities (~1600 kg/m^3) when the density should be around 630 kg/m^3. I equilibrated a system of 5000 CO2 molecules with Nose-Hoover and Berendsen for 5 ns, then did a production run using Nose-Hoover and Parrinello-Rahman for another 5 ns using 1 fs time step. I have found that tau_p plays an important factor. A relaxed tau_p gives low density while a vigorous tau_p gives a high density and crashes using Parrinello-Rahman pressure coupling. Has anyone else encountered similar problems or have any advice? Here is a copy of my MDP file. Any advice is greatly appreciated. Thanks Jarrett Wise PhD Student Petroleum Engineering University of Wyoming title = CO2 Density Run integrator = md nsteps = 500 dt = 0.001 nstxout = 0 nstvout = 0 nstfout = 0 nstxtcout = 2 nstenergy = 100 nstlog = 1500 constraint_algorithm = lincs constraints = H-Bonds lincs_iter = 1 lincs_order = 4 cutoff-scheme = Verlet ns_type = grid nstlist = 100 rlist = 1.4 rcoulomb= 1.4 rvdw= 1.4 nstcalcenergy = 10 comm_mode = linear coulombtype = PME pme_order = 4 fourierspacing = 0.16 tcoupl = Nose-Hoover tc-grps = System tau_t = 5.0 ref_t = 320 pcoupl = Parrinello-Rahman pcoupltype = isotropic tau_p = 3.0 ref_p = 520.0 compressibility = 1.45e-4 pbc = xyz continuation= yes gen_vel = no gen_temp = 320 -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
Re: [gmx-users] Supercritical CO2
Only other suggestion I have is that 5 ns production runs are probably not long enough. Granted, we needed diffusion data, so I believe I had 50 to 100 ns long simulations, but even for the barostat to settle, it may be a good idea to have simulations significantly longer than 5 ns. Alex On Thu, Apr 30, 2015 at 3:14 PM, Alex nedoma...@gmail.com wrote: Unfortunately, I had a pretty terrible experience trying to simulate fuels, fuel mixtures, and CO2 dissolved in fuels with Gromacs (using OPLS-AA) at high pressures and/or temperatures. It is my understanding that at the moment this is isn't the area where these forcefields are seriously tested. I have no results for other forcefields, but for OPLS-AA it was unpublishable and came out as an internal note at the institution. If curious, see http://nvlpubs.nist.gov/nistpubs/TechnicalNotes/NIST.TN.1805.pdf In fact, if you get any results that correspond to experimental data, I'll really appreciate it if you let me know. :) Alex On Thu, Apr 30, 2015 at 2:04 PM, Jarrett Lee Wise jwi...@uwyo.edu wrote: Hello all- I am trying to run simulations containing supercritical CO2 and oil asphaltenes. I have tested various force fields (TraPPE flex, TraPPe rigid, Charmm, EPM2, Cygan, and Zhang), but all of my results give either extremely low densities (~200 kg/m^3) or extremely high densities (~1600 kg/m^3) when the density should be around 630 kg/m^3. I equilibrated a system of 5000 CO2 molecules with Nose-Hoover and Berendsen for 5 ns, then did a production run using Nose-Hoover and Parrinello-Rahman for another 5 ns using 1 fs time step. I have found that tau_p plays an important factor. A relaxed tau_p gives low density while a vigorous tau_p gives a high density and crashes using Parrinello-Rahman pressure coupling. Has anyone else encountered similar problems or have any advice? Here is a copy of my MDP file. Any advice is greatly appreciated. Thanks Jarrett Wise PhD Student Petroleum Engineering University of Wyoming title = CO2 Density Run integrator = md nsteps = 500 dt = 0.001 nstxout = 0 nstvout = 0 nstfout = 0 nstxtcout = 2 nstenergy = 100 nstlog = 1500 constraint_algorithm = lincs constraints = H-Bonds lincs_iter = 1 lincs_order = 4 cutoff-scheme = Verlet ns_type = grid nstlist = 100 rlist = 1.4 rcoulomb= 1.4 rvdw= 1.4 nstcalcenergy = 10 comm_mode = linear coulombtype = PME pme_order = 4 fourierspacing = 0.16 tcoupl = Nose-Hoover tc-grps = System tau_t = 5.0 ref_t = 320 pcoupl = Parrinello-Rahman pcoupltype = isotropic tau_p = 3.0 ref_p = 520.0 compressibility = 1.45e-4 pbc = xyz continuation= yes gen_vel = no gen_temp = 320 -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Supercritical CO2
Yup, that would likely be pretty publishable. It kind of goes beyond just CO2. If you look at the report I mentioned above, there were also benzene and heptane, all wrong. I would then think that many other compounds relevant to fuel technology are also inadequately described. I'd say that a lot of people would appreciate a reasonable model that allows for modeling of hydrocarbon fuels and their mixtures in realistic conditions (high T/P). I myself am not immediately interested in working on this, but if you are, please let me know and I'll put you in touch with people who will provide the experimental data of highest quality for validation. This is sort of what my group and a few adjacent groups specialize in. On Fri, May 1, 2015 at 1:15 PM, V.V.Chaban vvcha...@gmail.com wrote: If no good model of CO2 still exists, we can publish a right one. It is very simple. I assume that C-O bond polarity is significantly different in the supercritical state and in the gas state. On Fri, May 1, 2015 at 3:30 PM, Alex nedoma...@gmail.com wrote: I wasn't talking about the initial post. I was talking about the fact that in _very stable_ states the same numerical simulation with a supposedly well-established set of interactions produces garbage. VVC The point is that any metastable state in numerical simulations VVC requires fine-tuned Hamiltonians. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Supercritical CO2
I wasn't talking about the initial post. I was talking about the fact that in _very stable_ states the same numerical simulation with a supposedly well-established set of interactions produces garbage. VVC The point is that any metastable state in numerical simulations VVC requires fine-tuned Hamiltonians. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Supercritical CO2
A cursory search yields: http://www.sciencedirect.com/science/article/pii/0009261496003971# http://www.sciencedirect.com/science/article/pii/S0009261403006407 http://link.springer.com/article/10.1007%2Fs11814-011-0248-5 The first one has some parameters, maybe trying them would be a start. Alex On Fri, May 1, 2015 at 3:40 PM, Jarrett Lee Wise jwi...@uwyo.edu wrote: Does anyone know if any of the typical CO2 force fields are parameritized for supercritical CO2? But if anyone does publish (or have) a good model, I would very much appreciate the parameters. Thanks again for all the input. Jarrett -Original Message- From: gromacs.org_gmx-users-boun...@maillist.sys.kth.se [mailto: gromacs.org_gmx-users-boun...@maillist.sys.kth.se] On Behalf Of Alex Sent: Friday, May 01, 2015 1:32 PM To: Vitaly Chaban Cc: gmx-users Subject: Re: [gmx-users] Supercritical CO2 Yup, that would likely be pretty publishable. It kind of goes beyond just CO2. If you look at the report I mentioned above, there were also benzene and heptane, all wrong. I would then think that many other compounds relevant to fuel technology are also inadequately described. I'd say that a lot of people would appreciate a reasonable model that allows for modeling of hydrocarbon fuels and their mixtures in realistic conditions (high T/P). I myself am not immediately interested in working on this, but if you are, please let me know and I'll put you in touch with people who will provide the experimental data of highest quality for validation. This is sort of what my group and a few adjacent groups specialize in. On Fri, May 1, 2015 at 1:15 PM, V.V.Chaban vvcha...@gmail.com wrote: If no good model of CO2 still exists, we can publish a right one. It is very simple. I assume that C-O bond polarity is significantly different in the supercritical state and in the gas state. On Fri, May 1, 2015 at 3:30 PM, Alex nedoma...@gmail.com wrote: I wasn't talking about the initial post. I was talking about the fact that in _very stable_ states the same numerical simulation with a supposedly well-established set of interactions produces garbage. VVC The point is that any metastable state in numerical simulations VVC requires fine-tuned Hamiltonians. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Epsilon-rf
That's a pretty poorly formed question, including the part where you actually calculate the epsilon-rf value. Can you please try again with a description of your system? Alex MK Dear all, MK I use reaction-field-zero as the coulombtype for my system. I wonder how I MK should choose the right epsilon-rf. How can I calculate the dielectric MK constant in the system? MK Thank you. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Epsilon-rf
With reaction-field-zero, the only option is epsilon-rf=0. That aside, if the nanotube is small, i would use cut-off electrostatics with coulomb-cutoff comparable to the length of the tube. If the tube is long, I would use standard pme. I would be extremely careful about using truncated electrostatics in general. Given the above, I am not sure I understand what's there to calculate, unless you mean running a simulation and _obtaining_ a dielectric property as a result. In that case, given you system, it is a completely different discussion. Alex MK Not a poorly formed but a general question! MK You are right about reaction-field-zero. reaction-field leads to MK bad energy conservations. So, I decided to use reaction-field-zero MK but as I have some charges I don't know what to do with them. I MK have a CNT in which there are some ions. I want to calculate the MK epsilon-rf for it. I don't know if epsilon-rf belongs to the whole MK system and if so, how I should calculate it. MK thanks! MK On Tue, Apr 28, 2015 at 10:50 AM, Alex nedoma...@gmail.com wrote: MK That's a pretty poorly formed question, including the part where you MK actually "calculate" the epsilon-rf value. Can you please try again with MK a description of your system? MK Alex MK Dear all, MK I use reaction-field-zero as the coulombtype for my system. I wonder how I MK should choose the right epsilon-rf. How can I calculate the dielectric MK constant in the system? MK Thank you. -- Best regards, Alex mailto:nedoma...@gmail.com -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Epsilon-rf
Also, reaction-field-zero, as described here http://manual.gromacs.org/online/mdp_opt.html which _only_ works with epsilon-rf=0. Any particular reason for doing this to your electrostatics? Alex MK Dear all, MK I use reaction-field-zero as the coulombtype for my system. I wonder how I MK should choose the right epsilon-rf. How can I calculate the dielectric MK constant in the system? MK Thank you. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Epsilon-rf
Yup, that should be physically sound. Alex Dear Alex, I have another question related to coulombtype selection. I dont want to use pbc in my system so, PME and Ewald are not good choices. If I use Reaction-field-zero and define a rcoulomb equal to my system size ( since epsilon-rf is 0 out of cutoff size) would it do the same work as finding epsilon-rf in generalized field reaction? THANKS! On Tue, Apr 28, 2015 at 10:22 PM, Maryam Kowsar maryam.kow...@gmail.com wrote: If I use PME there's no need to find epsilon-rf, but I have to change some other parameters. I will try it. Thank you very much Alex! On Tue, Apr 28, 2015 at 11:31 AM, Alex nedoma...@gmail.com wrote: With reaction-field-zero, the only option is epsilon-rf=0. That aside, if the nanotube is small, i would use cut-off electrostatics with coulomb-cutoff comparable to the length of the tube. If the tube is long, I would use standard pme. I would be extremely careful about using truncated electrostatics in general. Given the above, I am not sure I understand what's there to calculate, unless you mean running a simulation and _obtaining_ a dielectric property as a result. In that case, given you system, it is a completely different discussion. Alex MK Not a poorly formed but a general question! MK You are right about reaction-field-zero. reaction-field leads to MK bad energy conservations. So, I decided to use reaction-field-zero MK but as I have some charges I don't know what to do with them. I MK have a CNT in which there are some ions. I want to calculate the MK epsilon-rf for it. I don't know if epsilon-rf belongs to the whole MK system and if so, how I should calculate it. MK thanks! MK On Tue, Apr 28, 2015 at 10:50 AM, Alex nedoma...@gmail.com wrote: MK That's a pretty poorly formed question, including the part where you MK actually "calculate" the epsilon-rf value. Can you please try again with MK a description of your system? MK Alex MK Dear all, MK I use reaction-field-zero as the coulombtype for my system. I wonder how I MK should choose the right epsilon-rf. How can I calculate the dielectric MK constant in the system? MK Thank you. -- Best regards, Alex mailto:nedoma...@gmail.com -- Gromacs Users mailing list * Please search the archive athttp://www.gromacs.org/Support/Mailing_Lists/GMX-Users_Listbefore posting! * Can't post? Readhttp://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-usersor send a mail togmx-users-requ...@gromacs.org. -- Best regards, Alex mailto:nedoma...@gmail.com -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] question
Hi, I am not sure anyone will be able to help, because this isn't really about Gromacs. Can you please clarify who they are and what exactly you mean by perpendicular when talking about surfactants? Also, what are those surfactants? Anyway, a reference to a representative paper and some instructions on where to quickly see what you're talking about would be nice. Once again, this isn't something Gromacs-specific. Alex On May 15, 2015 9:17 AM, Golnaz Roudsari golnaz.c...@gmail.com wrote: Dear gmx-user I want to simulate graphene and surfactants in water. In most of the paper they put surfactants perpendicular on the graphene sheet. I have problem in initial configuration. I do not know how They put them perpendicular on the graphene sheet. I can not put them random in my box because I do not have a perfect system for more than 50 ns MD run. Thank you in advance Golnaz Vahabpour -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] gromacs.org_gmx-users Digest, Vol 133, Issue 87
Golnaz, I think you're creating a bit of a mess by replying to the digest. If you don't mind, please see my answer directly to your specific question, where I will also quote your message from below. Alex GR Thank you Alex GR First of all it is better to say that I am a beginner. I follow two papers GR for my project. The authers are ''Tumala'' and ''Samalkorpi'' . GR My surfactants are mixture of CTAB and SDS. In these two papers surfactants GR were put perpendicular on the graphene sheet. because if I put surfactants GR random in the box it will take a alot of time for surfactants to transfer GR to the graphene sheet. The simulation method in the papers is GROMACS but GR authors did not mention how they loacted surfactants on the graphene sheet. GR I have written the link of the papers. I knew that it is not specifically GR Gromacs but I would like to know with which software I can make this GR change on configuration. GR http://pubs.acs.org/doi/abs/10.1021/jp077636y GR http://pubs.acs.org/doi/abs/10.1021/jp077678m GR On Fri, May 15, 2015 at 7:14 PM, GR gromacs.org_gmx-users-requ...@maillist.sys.kth.se wrote: Send gromacs.org_gmx-users mailing list submissions to gromacs.org_gmx-users@maillist.sys.kth.se To subscribe or unsubscribe via the World Wide Web, visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or, via email, send a message with subject or body 'help' to gromacs.org_gmx-users-requ...@maillist.sys.kth.se You can reach the person managing the list at gromacs.org_gmx-users-ow...@maillist.sys.kth.se When replying, please edit your Subject line so it is more specific than Re: Contents of gromacs.org_gmx-users digest... Today's Topics: 1. Re: energy minimization problem (Ming Tang) 2. Re: energy minimization problem (Justin Lemkul) 3. question (Golnaz Roudsari) 4. Re: question (Alex) 5. Re: question (abhijit Kayal) -- Message: 1 Date: Fri, 15 May 2015 13:32:04 + From: Ming Tang m21.t...@qut.edu.au To: gmx-us...@gromacs.org gmx-us...@gromacs.org Subject: Re: [gmx-users] energy minimization problem Message-ID: dv888t2enjj742h6jn4fglmx.1431696517...@email.android.com Content-Type: text/plain; charset=us-ascii Reducing time helps it run longer. But 0.02ps-0.04ps is believed to be reasonable according to gromacs website. So ijust kept trying 0.02ps Sent from my Huawei Mobile Justin Lemkul jalem...@vt.edu wrote: On 5/14/15 10:02 PM, Ming Tang wrote: Hi Justin Here is the .mdp file with position restraints, which I downloaded from the martini tutorial website. It works well with the tutorial and my small triple helix, and the grompp process does not give any warnings. Please help to have a look. Many thanks to you. define = -DPOSRES integrator = md dt = 0.02 Have you tried simply reducing the time step? 20 fs may be reasonable for a well equilibrated system, but if the run is dying early, then this is the first thing I would look at. -Justin nsteps = 5 nstxout = 0 nstvout = 0 nstlog = 1000 nstxtcout= 1000 xtc-precision= 10 cutoff-scheme=verlet coulombtype = reaction-field-zero coulomb-modifier = potential-shift-verlet rcoulomb-switch = 0.9 rcoulomb = 1.4 epsilon_r= 15 vdw-modifier = force-switch rvdw-switch = 0.9 rvdw = 1.4 tcoupl = v-rescale tc-grps = Protein W tau-t= 1.0 1.0 ref-t= 310 310 compressibility = 3e-4 ref-p= 1.0 refcoord_scaling = all pbc = xyz -Original Message- From: gromacs.org_gmx-users-boun...@maillist.sys.kth.se [mailto: gromacs.org_gmx-users-boun...@maillist.sys.kth.se] On Behalf Of Justin Lemkul Sent: Friday, 15 May 2015 11:50 AM To: gmx-us...@gromacs.org Subject: Re: [gmx-users] energy minimization problem On 5/14/15 9:47 PM, Ming Tang wrote: Dear Justin, I saw it in the terminal. Potential Energy = -6.3682131e+05 Maximum force = 1.8876091e+02 on atom 1147 Norm of force = 3.1155257e+00 But, how can I further minimize my protein? Minimization is not your problem. Post a full .mdp file of the run that is failing. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search
Re: [gmx-users] question
Okay, I looked at the first paper. As Abhijit said, you can manipulate any part of your input coordinates by using something like packmol. In addition, pymol and GMX itself (editconf) provides means for translation and rotation of your choosing. I think VMD has this functionality as well, although I only use it as a viewer. There's little to add at this point: learn the tool(s) of your choice and use them to manipulate your coordinates. Alex GR Thank you Alex GR First of all it is better to say that I am a beginner. I follow two papers GR for my project. The authers are ''Tumala'' and ''Samalkorpi'' . GR My surfactants are mixture of CTAB and SDS. In these two papers surfactants GR were put perpendicular on the graphene sheet. because if I put surfactants GR random in the box it will take a alot of time for surfactants to transfer GR to the graphene sheet. The simulation method in the papers is GROMACS but GR authors did not mention how they loacted surfactants on the graphene sheet. GR I have written the link of the papers. I knew that it is not specifically GR Gromacs but I would like to know with which software I can make this GR change on configuration. GR http://pubs.acs.org/doi/abs/10.1021/jp077636y GR http://pubs.acs.org/doi/abs/10.1021/jp077678m Hi, I am not sure anyone will be able to help, because this isn't really about Gromacs. Can you please clarify who "they" are and what exactly you mean by "perpendicular" when talking about surfactants? Also, what are those surfactants? Anyway, a reference to a representative paper and some instructions on where to quickly see what you're talking about would be nice. Once again, this isn't something Gromacs-specific. Alex On May 15, 2015 9:17 AM, "Golnaz Roudsari" golnaz.c...@gmail.com wrote: Dear gmx-user I want to simulate graphene and surfactants in water. In most of the paper they put surfactants perpendicular on the graphene sheet. I have problem in initial configuration. I do not know how They put them perpendicular on the graphene sheet. I can not put them random in my box because I do not have a perfect system for more than 50 ns MD run. Thank you in advance Golnaz Vahabpour -- Gromacs Users mailing list * Please search the archive athttp://www.gromacs.org/Support/Mailing_Lists/GMX-Users_Listbefore posting! * Can't post? Readhttp://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-usersor send a mail togmx-users-requ...@gromacs.org. -- Best regards, Alex mailto:nedoma...@gmail.com -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Carbon Nanotube `Y-Junctions'
The only software I am aware of that can do this out of the box is Nanotube Modeler: http://jcrystal.com/products/wincnt/ It is a paid product full of less than obvious bugs, but for the most part it works. The free demo will let you see the structures it generates, but you won't be able to export anything. Alex On Thu, May 14, 2015 at 10:35 PM, saeed poorasad s_poora...@yahoo.com wrote: Dear Gromacs users , Greetings .I am looking for XYZ or PDB file of Carbon nanotube ' Y-junction' . I will be grateful if you can let me know how i can find or make it . Thanks in advance .Best ,Saeed. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Topology file for Fullerene with OPLS_AA force field
For fullerenes, follow http://www.webcitation.org/66u2xJJ3O and use x2top to build the topology based on OPLSAA. Alex On Wed, Apr 15, 2015 at 12:51 AM, saeed poorasad s_poora...@yahoo.com wrote: Hi Gromacs users , I want to have a system with one C60 and about 1000 water molecules .I tried to use pdb2gmx on C60.pdb but I got fatal error , I think the problem is because of my PDB file .I want to define C-delta atom of the aromatic tryptophan amino acid for C60 carbons and i want to use oplsaa force field. this is a part of my PDB file : HETATM1 CD1 TRP 0 3.341 0.669 0.088 1.00 0.00 CHETATM2 CD1 TRP 0 3.266 -0.548 -0.502 1.00 0.00 CHETATM3 CD1 TRP 0 2.990 -1.662 0.229 1.00 0.00 CHETATM4 CD1 TRP 0 2.750 -0.402 -1.741 1.00 0.00 C I have no experience with GROMACS , so I will be grateful if you can help me . Best ,Saeed. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] POPE lipid from Tieleman's web site, the box of xy dimension will shrink during the simulation
Though I doubt it will make a difference, it may be useful to put a space between the values in ref_p= 1.01.0; reference pressure, x-y, z (in bar) compressibility = 4.5e-54.5e-5; isothermal compressibility, bar^-1 tau_t= 0.50.5;0.5; time constant, in ps but in any case, this could be poor pre-equilibration prior to dynamic simulations. Also, you have a relatively loose barostat in conjunction with a loose thermostat (0.5 ps), so the initial shockwaves linger in the system. Try the default constants of tau_p=2.0 and tau_t = 0.1. Nose-Hoover thermostat (algorithmically oscillatory) does not help with long ref_t either, maybe try v-rescale. Alex n Dear all, nI am new to lipid simulation. I have downloaded the pope.pdb, n pope.itp, lipid.itp from Tieleman's web site n (http://wcm.ucalgary.ca/tieleman/downloads). And I have made a n directory named gromos53a6_lipid.ff followed Justin's n tutorial n (http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin/gmx-tutorials/membrane_protein/index.html). n Then I use pope.pdp to run a pure 340 POPE lipid MD at 310K. But I n find that the box of xy dimension will shrink from n 9.6nm*9.5nm to 9.0nm*8.9nm after 20 ns, then it fluctuates at that n value till 200ns. The box of z dimension will increase a little to n hold the volume constant. This will lead to the area per lipid decreased from 0.54 nm^2 to n 0.47nm^2. At the same time, I find the order parameter of lipid n tail (Scd) is too big (above 0.3). Does any one know where the problem is? n When I use dppc128.pdb from Tieleman's web site to run a pure n DPPC lipid MDat 323K, the box will not shrink. And the area per n lipid of DPPC is in accordance with literature, about 0.62nm^2. n The water model is SPC. My mdp file is also from Justin's n tutorial. Followed is my mdp file of POPE: n title= KALP15-DPPC Production MD n ; Run parameters n integrator= md; leap-frog integrator n nsteps= 1; 2 * 50 = 1000 ps (1 ns) n dt= 0.002; 2 fs n ; Output control n nstxout= 0; save coordinates every 2 ps n nstvout= 0; save velocities every 2 ps n nstxtcout= 1000; xtc compressed trajectory output every 2 ps n nstenergy= 1000; save energies every 2 ps n nstlog= 1000; update log file every 2 ps n ; Bond parameters n continuation= yes; Restarting after NPT n constraint_algorithm = lincs; holonomic constraints n constraints= all-bonds; all bonds (even heavy atom-H bonds) constrained n lincs_iter= 1; accuracy of LINCS n lincs_order= 4; also related to accuracy n ; Neighborsearching n ns_type= grid; search neighboring grid cels n nstlist= 5; 10 fs n rlist= 1.2; short-range neighborlist cutoff (in nm) n rcoulomb= 1.2; short-range electrostatic cutoff (in nm) n rvdw= 1.2; short-range van der Waals cutoff (in nm) n ; Electrostatics n coulombtype= PME; Particle Mesh Ewald for long-range electrostatics n pme_order= 4; cubic interpolation n fourierspacing= 0.16; grid spacing for FFT n ; Temperature coupling is on n tcoupl= Nose-Hoover; More accurate thermostat n tc-grps= POPE SOL;SOL_CL; three coupling groups - more accurate n tau_t= 0.50.5;0.5; time constant, in ps n ref_t= 310 310;323; reference temperature, one for each group, in K n ; Pressure coupling is on n pcoupl= Parrinello-Rahman; Pressure coupling on in NPT n pcoupltype= semiisotropic; uniform scaling of x-y box vectors, independent z n tau_p= 2.0; time constant, in ps n ref_p= 1.01.0; reference pressure, x-y, z (in bar) n compressibility = 4.5e-54.5e-5; isothermal compressibility, bar^-1 n ; Periodic boundary conditions n pbc= xyz; 3-D PBC n ; Dispersion correction n DispCorr= EnerPres; account for cut-off vdW scheme n ; Velocity generation n gen_vel= no; Velocity generation is off n ; COM motion removal n ; These options remove motion of the protein/bilayer relative to the solvent/ions n nstcomm = 1 n comm-mode = Linear n comm-grps = POPE SOL n ; Scale COM of reference coordinates (when postion restraint) n ;refcoord_scaling = com n And my topol file: n #include gromos53a6_lipid.ff/forcefield.itp n #include pope.itp n ; Include water topology n #include gromos53a6_lipid.ff/spc.itp n ; System specifications n [ system ] n 340-Lipid POPE and CNP n [ molecules ] n ; molecule name nr. n POPE 340 n SOL 6729 n Thanks in advance, n Nie Xuechuan n Shanghai Institute of Applied Physics, Chinese Academy of Sciences -- Best regards, Alexmailto:nedoma...@gmail.com -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] How to get the .gro file for MgO surface
You probably mean the topology, which you will have to build yourself, likely at the forcefield level. Is there any particular reason for trying to simulate metal oxide with GMX? There are much better tools suited for this, e.g. LAMMPS with ReaxFF. Alex z Dear you, z Does anyone know how to get the .gro file for MgO z surface? The pdb2gmx and Prodrg could not do that. z Best regards z Hongping -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
[gmx-users] R-B dihedrals
I have a set of dihedrals described by the R-B potential, with a minimum at 90 degrees, and minimization indeed yielded angles very close to that. At some point, I decided to check if I can rotate the groups to another angle, say, 45 degrees. Simple enough, so I changed the expression. Energy minimization shows no difference -- still 90 deg. I turned off vdw for all atoms in the vicinity, set all charges to zero -- no difference. Hiked up the dihedral constant by three(!) orders of magnitude -- nothing. Commenting out these dihedrals immediately causes grompp errors, so they are definitely seen by the system (I tracked the errors to the atomtypes). The dihedral type (3) is listed correctly in the topology. A David Lynch mystery. Any ideas? Thanks, Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] R-B dihedrals
Absolutely nothing in between, this is (cos(45)-cos(phi))^2, two minima at 45 and 315 and a peak around 180. Checked if I converted to R-B coefficients correctly -- same thing. At 90/270 clean decrease to 45/315. Alex On Tue, Apr 14, 2015 at 3:16 PM, Justin Lemkul jalem...@vt.edu wrote: On 4/14/15 5:12 PM, Alex wrote: Tried both, the (now enormous) parameters are read in correctly, nothing in output mdp that catches my eye. No constraints used in the system, explicitly set in mdp. Plotted the potential at 45/315 deg minimum -- definitely nonzero at 90 degrees (starting config). This has got to be the strangest GMX issue i've ever had. What does a 1-D energy surface look like for that dihedral? You've got a dihedral at 90 in the coordinates and you're trying to shift it to 45? If there's a massive barrier in between, EM won't necessarily do that for you. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/ Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] R-B dihedrals
Mark, we just figured it out, thanks. :) And yes, you are definitely onto something: two identical dihedrals pulling in opposite directions. This is what happens when people from solid state show up to use GMX. :) Alex On Tue, Apr 14, 2015 at 5:20 PM, Mark Abraham mark.j.abra...@gmail.com wrote: Can you shift it to 40 degrees? Mark On Tue, Apr 14, 2015 at 11:26 PM, Alex nedoma...@gmail.com wrote: Absolutely nothing in between, this is (cos(45)-cos(phi))^2, two minima at 45 and 315 and a peak around 180. Checked if I converted to R-B coefficients correctly -- same thing. At 90/270 clean decrease to 45/315. Alex On Tue, Apr 14, 2015 at 3:16 PM, Justin Lemkul jalem...@vt.edu wrote: On 4/14/15 5:12 PM, Alex wrote: Tried both, the (now enormous) parameters are read in correctly, nothing in output mdp that catches my eye. No constraints used in the system, explicitly set in mdp. Plotted the potential at 45/315 deg minimum -- definitely nonzero at 90 degrees (starting config). This has got to be the strangest GMX issue i've ever had. What does a 1-D energy surface look like for that dihedral? You've got a dihedral at 90 in the coordinates and you're trying to shift it to 45? If there's a massive barrier in between, EM won't necessarily do that for you. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/ Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] GMX-compatible DNA coordinates
Tsjerk, I can only describe the project in a private communication, as this is a completely new idea (potentially high-risk from the experimental standpoint) in a very competitive field. If you're interested, please let me know, and I'll email you directly. Other than that, there is not much I can reveal on a public forum. I hope you'll understand. ;) Alex On Fri, Apr 3, 2015 at 1:33 PM, Tsjerk Wassenaar tsje...@gmail.com wrote: Hi Alex, I've done it before, but I can't do it from the top of my head without seeing the input and checking what Gromacs wants. It's not hard work, it's not a lot of work, but it is work, and more than we'll usually consider doing for answering a post on the mailing list. On the other hand, you can always try to raise interest in a collaboration, which is a completely different matter :) Cheers, Tsjerk On Fri, Apr 3, 2015 at 8:21 PM, Alex nedoma...@gmail.com wrote: Hi Tsjerk, You're right. It's just that the last time I used a sed or awk script of my own composition was in college, which is ~13 years ago. :)) So, this is still an undertaking for someone like me. But yes, what you said is well taken. Lack of knowledge/desire to rename atoms in text files shouldn't hinder research. I do think that having a person such as yourself of Justin on my team to implement what I have in mind would speed things up tremendously, but noone seems to be interested, hint-hint. ;) Thanks, Alex Hi Alex, Writing a sed oneliner is not the same as writing a renaming script. I commonly use them to process PDB files in some way and it's surprising how much you can do without scripting. Cheers, Tsjerk On Apr 3, 2015 7:13 PM, Alex nedoma...@gmail.com wrote: Gentlemen, Building a script to rename the atom is obviously an option. I was just wondering if I missed something, e.g. a VMD plugin that could do it quickly. As far as google searching, I have done that, and I have found some things (like the make-NA server, or this: http://smog-server.org/cgi-bin/GenTopGro.pl). This is still not quite what I was aiming for. Hence, my initial question. I didn't mean to be snippy, but I did try searching, believe me. On the other hand, I assume that you are much better versed in the problem at hand, so then maybe you'd be able to appreciate info I simply ignored. Obviously, I welcome anything you can suggest. If this boils down to me writing a renaming script, so be it. :) Thanks, Alex On 03/04/2015 5:53 pm, Mark Abraham mark.j.abra...@gmail.com wrote: Hi, I tried that search myself and found some things that looked potentially useful to you, so I suggested that search. Either you might find something useful there, or the dialogue could progress about what you need. Requiring perfect input for pdb2gmx is a fairly stringent expectation. i.e. The use of sed or similar tools is normal, as Tsjerk says also. There are too many options, and nobody has built the Babel fish of MD input! Mark Mark On 03/04/2015 3:40 pm, Alex nedoma...@gmail.com wrote: Mark, Not to troll, but why did you assume that I didn't do my due diligence before posting here? Because, you see, you could have left that second part of your reply (which only confirms the legitimacy of my initial question), and it would make it less insulting. Alex Hi, Googling things like gromacs DNA structure is a good bet. There are no fully standardised naming schemes or workflows, so you will likely not have something do everything for you. Mark On 03/04/2015 7:03 am, Alex nedoma...@gmail.com wrote: I'm fully aware that this is a forum for discussing Gromacs, but there's a bit of a frustrating situation that involves using it. I need a linear ssDNA chain, sequence of my choosing, anything that will work with, say, AMBER99SB-ILDN. Not trying to find any funny forcefields for it, everything as is. You suggested AMBER in the past, so AMBER it is. ChemOffice produced a PDB with incompatible atom types. make-na server (http://structure.usc.edu/make-na/server.html) also produced something pdb2gmx hates. Is there absolutely any way to get a PDB that will produce a usable topology out of the box with an established potential implemented in Gromacs? Anything that will not involve manual modification of PDBs or writing text-processing scripts will make me very happy. I will really appreciate any suggestions. Thank you, Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org
Re: [gmx-users] Gromacs on Nvidia 830M
Of course, it's not CP2K, which will murder your laptop. :) Not for long production simulations, but great for prototyping and shorter tests. Gromacs under Linux is a treat. Alex BP Dear users, BP I perfectly understand that gromacs is not for laptops. However, I was BP wondering about gromacs performing on an i3 laptop with Nvidia 830M with BP 4Gb memory. Has anyone tried? BP Regards, BP BP. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
[gmx-users] grompp: unknown bond_atomtype
I am working with a local copy of oplsaa.ff where I have defined a set of custom atom types in atomname2type.n2t and added the bond and angle parameters into ffbonded. x2top creates the topology without issues, but grompp reports a fatal error (see subject). The particular type it complains about is actually opls_335. Gromacs v. 4.6-beta3. Any ideas? Thank you, Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] The largest charge group contains [GIANT CRAPTON] of atoms
Ah, now I see, thanks. Yes, this entirely follows the paradigm of biomolecules. I still insist that Gromacs was a better choice for my project than LAMMPS. :) On Thu, Apr 2, 2015 at 12:31 PM, Justin Lemkul jalem...@vt.edu wrote: On 4/2/15 2:28 PM, Alex wrote: Verlet option did remove that error, thank you. Indeed, there is a large number of atoms with identical cgnr, no idea why. Does it have to do with the order of atoms in your PDB? No, it has to do with the definition of a charge group. In the x2top.c function set_cgnr(), the charge group index is only incremented once an integral charge has been accumulated, per convention. Probably in your case, it takes a lot of atoms before that criterion is satisfied. -Justin On Thu, Apr 2, 2015 at 10:22 AM, Justin Lemkul jalem...@vt.edu wrote: On 4/1/15 7:42 PM, Alex wrote: I have a custom-FF (OPLSAA based) system with ~8500 atoms, which is a rectangular graphene sheet with a hole and six pieces of small organics covalently bound at the hole edge. About 95% of all atoms are electrically neutral, with the exception of the functional groups mentioned above and a total of 24 graphene edge carbons (-0.06e), each passivated by a hydrogen (+0.06e). All charges have been carefully set (actually, from our own DFT calculations) to make the system electrically neutral. The charge values only have two decimals, set in ffnonbonded. Two issues: 1. x2top yields a total charge of something e-6. I initially ignored it, but now I think it may be indicative of a actual problem, because: No, that magnitude of charge is not actually a problem. 2. grompp produces a fatal error (see subject) with the number of atoms in the group ~half the size of the system. Because I am not using anything out of the box (coordinates, FF), I'm really perplexed here. I checked the manual, and it makes sense that a charge group is a way to lump atoms into locally neutral regions. And that's sort of exactly what I have, at least from the standpoint of geometry. Definitely not worse than a peptide chain common for GMX. The cgnr column will tell you what's going on. Charge groups are irrelevant if using the Verlet cutoff scheme. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/ Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/ Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] The largest charge group contains [GIANT CRAPTON] of atoms
Verlet option did remove that error, thank you. Indeed, there is a large number of atoms with identical cgnr, no idea why. Does it have to do with the order of atoms in your PDB? On Thu, Apr 2, 2015 at 10:22 AM, Justin Lemkul jalem...@vt.edu wrote: On 4/1/15 7:42 PM, Alex wrote: I have a custom-FF (OPLSAA based) system with ~8500 atoms, which is a rectangular graphene sheet with a hole and six pieces of small organics covalently bound at the hole edge. About 95% of all atoms are electrically neutral, with the exception of the functional groups mentioned above and a total of 24 graphene edge carbons (-0.06e), each passivated by a hydrogen (+0.06e). All charges have been carefully set (actually, from our own DFT calculations) to make the system electrically neutral. The charge values only have two decimals, set in ffnonbonded. Two issues: 1. x2top yields a total charge of something e-6. I initially ignored it, but now I think it may be indicative of a actual problem, because: No, that magnitude of charge is not actually a problem. 2. grompp produces a fatal error (see subject) with the number of atoms in the group ~half the size of the system. Because I am not using anything out of the box (coordinates, FF), I'm really perplexed here. I checked the manual, and it makes sense that a charge group is a way to lump atoms into locally neutral regions. And that's sort of exactly what I have, at least from the standpoint of geometry. Definitely not worse than a peptide chain common for GMX. The cgnr column will tell you what's going on. Charge groups are irrelevant if using the Verlet cutoff scheme. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/ Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Setting custom atoms in FF
I am not mixing forcefield types, but I am indeed mixing angle types. I suppose I will adjust my constants so the local derivatives near equilibrium roughly correspond to what I need while sticking to the rtp default, I can't see another way around. I have tremendous respect for you guys, but I too developed a ton of code for atomistic methods and I do not see the rationale for imposing this rigidity. I would understand it, if the code used mathematically cumbersome potentials, but with all these polynomial springs, the forces could be done in a switchable manner in a single sweep... Thanks for clarifying this, Justin. Kind of a crucial point. Alex On Wed, Apr 1, 2015 at 2:26 PM, Justin Lemkul jalem...@vt.edu wrote: On 4/1/15 4:24 PM, Alex wrote: That is the issue. I AM using -noparam, and x2top is reverting to an incorrect bond type. Which then sets off errors in grompp, because indeed it looks up the bond in ffbonded and sees nothing for the bond type just produced by x2top. As soon as I manually modify bond types in topology, the problem is solved. The function type is read from the header of the .rtp. If you want to change the way the force field works, modify that. If you're mixing types, you're out of luck in terms of automation. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/ Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Setting custom atoms in FF
I will certainly consider it. In all honesty, however, the situation is due to the fact that I have graphene constants that have been calculated a certain way from the Brenner potential. The problem at hand can be solved without modifying GMX code by merely calculating proper constants for _all_ angle function types (within GMX convention). I recall posting a while back about properly parameterizing graphene, and heard mostly meh from the locals (I don't think you participated in that discussion). Alex On Wed, Apr 1, 2015 at 2:46 PM, Justin Lemkul jalem...@vt.edu wrote: On 4/1/15 4:37 PM, Alex wrote: I am not mixing forcefield types, but I am indeed mixing angle types. I suppose I will adjust my constants so the local derivatives near equilibrium roughly correspond to what I need while sticking to the rtp default, I can't see another way around. I have tremendous respect for you guys, but I too developed a ton of code for atomistic methods and I do not see the rationale for imposing this rigidity. I would understand it, if the code used mathematically It's a simple convention that works well for biomolecular systems, for which GROMACS is principally suited. It works well in nearly all cases, but clearly not here. If you want to file a feature request, we welcome feedback and input on redmine.gromacs.org, but the developers do what they have time for based on their interests. If you want to submit a patch to introduce new functionality, great! Please do; we need contributions from users for things like this. -Justin cumbersome potentials, but with all these polynomial springs, the forces could be done in a switchable manner in a single sweep... Thanks for clarifying this, Justin. Kind of a crucial point. Alex On Wed, Apr 1, 2015 at 2:26 PM, Justin Lemkul jalem...@vt.edu wrote: On 4/1/15 4:24 PM, Alex wrote: That is the issue. I AM using -noparam, and x2top is reverting to an incorrect bond type. Which then sets off errors in grompp, because indeed it looks up the bond in ffbonded and sees nothing for the bond type just produced by x2top. As soon as I manually modify bond types in topology, the problem is solved. The function type is read from the header of the .rtp. If you want to change the way the force field works, modify that. If you're mixing types, you're out of luck in terms of automation. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/ Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/ Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] grompp: unknown bond_atomtype
True, true. :) On Wed, Apr 1, 2015 at 2:29 PM, Justin Lemkul jalem...@vt.edu wrote: On 4/1/15 4:27 PM, Alex wrote: I think I solved it already, thanks. Entirely due to horrible mess created by the necessity to create custom atom types. I wish you guys could make the bond length tolerance an input to x2top, I really do. :) Well, I already showed you the function to modify to change it to your heart's desire ;) -Justin On Wed, Apr 1, 2015 at 2:18 PM, Justin Lemkul jalem...@vt.edu wrote: On 4/1/15 2:03 PM, Alex wrote: I am working with a local copy of oplsaa.ff where I have defined a set of custom atom types in atomname2type.n2t and added the bond and angle parameters into ffbonded. x2top creates the topology without issues, but grompp reports a fatal error (see subject). The particular type it complains about is actually opls_335. Gromacs v. 4.6-beta3. Any ideas? Don't use beta versions, especially ancient ones. The problem is related to a translated type; note that in ffnonbonded.itp the second column is an atom type that is used in ffbonded.itp. These are different from the nonbonded types, but since there are fewer bonded types to cover the parameters, these are used to remove redundancy. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/ Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/ Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] grompp: unknown bond_atomtype
I think I solved it already, thanks. Entirely due to horrible mess created by the necessity to create custom atom types. I wish you guys could make the bond length tolerance an input to x2top, I really do. :) On Wed, Apr 1, 2015 at 2:18 PM, Justin Lemkul jalem...@vt.edu wrote: On 4/1/15 2:03 PM, Alex wrote: I am working with a local copy of oplsaa.ff where I have defined a set of custom atom types in atomname2type.n2t and added the bond and angle parameters into ffbonded. x2top creates the topology without issues, but grompp reports a fatal error (see subject). The particular type it complains about is actually opls_335. Gromacs v. 4.6-beta3. Any ideas? Don't use beta versions, especially ancient ones. The problem is related to a translated type; note that in ffnonbonded.itp the second column is an atom type that is used in ffbonded.itp. These are different from the nonbonded types, but since there are fewer bonded types to cover the parameters, these are used to remove redundancy. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/ Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] GMX-compatible DNA coordinates
Mark, Not to troll, but why did you assume that I didn't do my due diligence before posting here? Because, you see, you could have left that second part of your reply (which only confirms the legitimacy of my initial question), and it would make it less insulting. Alex Hi, Googling things like "gromacs DNA structure" is a good bet. There are no fully standardised naming schemes or workflows, so you will likely not have something do everything for you. Mark On 03/04/2015 7:03 am, "Alex" nedoma...@gmail.com wrote: I'm fully aware that this is a forum for discussing Gromacs, but there's a bit of a frustrating situation that involves using it. I need a linear ssDNA chain, sequence of my choosing, anything that will work with, say, AMBER99SB-ILDN. Not trying to find any funny forcefields for it, everything as is. You suggested AMBER in the past, so AMBER it is. ChemOffice produced a PDB with incompatible atom types. make-na server (http://structure.usc.edu/make-na/server.html) also produced something pdb2gmx hates. Is there absolutely any way to get a PDB that will produce a usable topology out of the box with an established potential implemented in Gromacs? Anything that will not involve manual modification of PDBs or writing text-processing scripts will make me very happy. I will really appreciate any suggestions. Thank you, Alex -- Gromacs Users mailing list * Please search the archive athttp://www.gromacs.org/Support/Mailing_Lists/GMX-Users_Listbefore posting! * Can't post? Readhttp://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-usersor send a mail togmx-users-requ...@gromacs.org. -- Best regards, Alex mailto:nedoma...@gmail.com -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] cudaStreamSynchronize failed in cu_blockwait_nb
Additional info: The base system is a SuperMicro X8DTG-D motherboard with dual Xeon E5620 CPUs at 2.4 GHz, and 24 GB of DDR3 ECC RAM The GPUs are two NVidia Tesla C2075s Running under Ubuntu 14.04 LTS 64-bit, kernel 3.13.0-53, with CUDA 7.0, installed via the NVIDIA-provided .deb file, cuda-repo-ubuntu1404-7-0-local_7.0-28_amd64.deb, followed by the usual apt-get update/upgrade. Nvcc reports version 7.0.27. GMX version is 5.0.5, locally built, with auto-detected options, including GMX_GPU ON, and GMX_SIMD SSE4.1 On Mon, Jun 8, 2015 at 2:07 PM, Alex nedoma...@gmail.com wrote: Hi all, I am running a GPU-accelerated test (GMX 5.0.5) and, after about 60K steps of what seems to be a huge speed-up, I get the error from the subject with the following comment: unspecified launch failure. This is on an 8-core (16-thread) Intel CPU and two Tesla cards. The run command is: mdrun -ntomp 4 file directives I saw some previous messages on this error, but not sure there's a solution. Any thoughts? I will provide any additional info as needed. If there are additional options necessary to debug etc, could you provide the exact command line? Thanks, Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
[gmx-users] cudaStreamSynchronize failed in cu_blockwait_nb
Hi all, I am running a GPU-accelerated test (GMX 5.0.5) and, after about 60K steps of what seems to be a huge speed-up, I get the error from the subject with the following comment: unspecified launch failure. This is on an 8-core (16-thread) Intel CPU and two Tesla cards. The run command is: mdrun -ntomp 4 file directives I saw some previous messages on this error, but not sure there's a solution. Any thoughts? I will provide any additional info as needed. If there are additional options necessary to debug etc, could you provide the exact command line? Thanks, Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
[gmx-users] GPU acceleration
Hi all, We're about to start a bit of a toy project to test GPU-accelerated GMX. What I have here is my old workstation (Dell Precision T3500: Quad-core Xeon W3530 @ 2.8 GHz, 12 gigs of RAM) running Linux Mint 17.1. The plan is to get something like GeForce GTX 750 ( http://www.amazon.com/GeForce-750Ti-GDDR5-Graphics-GTX750TI-OC-2GD5/dp/B00IB9P1KG/ref=sr_1_3?ie=UTF8qid=1433529993sr=8-3keywords=Nvidia+GTX750 ) and maybe a new PSU. Questions: 1. With a card like this, how much acceleration (very approximately) can we expect in a standard simulation (protein+lipid+water, PME electrostatics)? 2. Should we put any extra considerations into cooling? We expect week-long runs, and I am not sure these cards have this type of hardcore gaming in mind. :) Any additional comments are highly appreciated. Please keep in mind that this is a toy project to utilize an old workstation. Thank you, Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] GPU acceleration
Thanks for the info. I tried with GMX_EMULATE_GPU=1 preceding the mdrun line, crashed after 100 steps. Is that the expected behavior? On Fri, Jun 5, 2015 at 2:46 PM, Szilárd Páll pall.szil...@gmail.com wrote: On Fri, Jun 5, 2015 at 10:39 PM, Alex nedoma...@gmail.com wrote: In that case, I am a bit lost, as I was sure GMXRC needed modification. :)). Since I never ran anything on GPU and have an mdrun line that looks like mdrun -nt 4 -s tpr -c final structure gmx mdrun -ntomp 4 No need to do domain decomposition. Even with CPU runs it'll be faster if you don't. what should be added to trigger the behavior you're suggesting? I understand that it will likely blow up the system, but for the heck of it... those environment variables I mentioned. Plus timer resetting, e.g. with -resethway if you want reliable timings. I am now looking at the stats and see that (NxN QSTab + LJ) and (NxN QSTab [F]) take up about 86% of all flops in my simulation. Which of those get offloaded? All of that will be offloaded. However, those are theoretical (useful) flops, the timing table on the bottom is more relevant. Set the GMX_CYCLE_SUBCOUNTER=ON breakdown of the Force. However, if you just follow the previous emulate+skip non-bondeds advise, you'll get a run that does the offload except that it skips calculating - which is useful because you have no GPU to calculate with. Thank you, Alex On Fri, Jun 5, 2015 at 2:26 PM, Szilárd Páll pall.szil...@gmail.com wrote: Hi Alex, Why do you think GMXRC needs to be changed? You simply need to set those env. vars to trigger the GPU code-path - and you will likely have to freeze the system, otherwise it will probably not run more than a few steps without seeing an explosion. Based on this you'll know how fast you can run with GPU offload, but you can also guesstimate how fast of a GPU you need (based on the PME+Bonded force calculation time + publicly available CUDA kernel performance). Cheers, -- Szilárd On Fri, Jun 5, 2015 at 10:13 PM, Alex nedoma...@gmail.com wrote: Hi Szilárd, Well, yes, it is an old CPU, but I am actually quite content with the ns/day performance it churns out on all four cores. I'm looking at http://manual.gromacs.org/documentation/5.1-beta1/user-guide/environment-variables.html and it is my understanding that GMXRC needs to be changed to include those flags, but no recompiling is required. Is that correct? Thanks, Alex On Fri, Jun 5, 2015 at 1:52 PM, Szilárd Páll pall.szil...@gmail.com wrote: Hi Alex, That's a quite old CPU so it's hard to tell how fast is it, you'll have to try. There are ways to guesstimate how fast does your GPU have to be to keep up with your CPU cores (hint search for GMX_EMULATE_GPU / GMX_NO_NONBONDED on the list). In any case, unless your CPU is so slow that it balances well with the 750Ti, I would strongly recommend that you get a GTX 960 which will cost just a bit more, but it will be 1.6-1.8x faster at computing non-bonded interactions. On Fri, Jun 5, 2015 at 9:08 PM, Alex nedoma...@gmail.com wrote: Hi all, We're about to start a bit of a toy project to test GPU-accelerated GMX. What I have here is my old workstation (Dell Precision T3500: Quad-core Xeon W3530 @ 2.8 GHz, 12 gigs of RAM) running Linux Mint 17.1. The plan is to get something like GeForce GTX 750 ( http://www.amazon.com/GeForce-750Ti-GDDR5-Graphics-GTX750TI-OC-2GD5/dp/B00IB9P1KG/ref=sr_1_3?ie=UTF8qid=1433529993sr=8-3keywords=Nvidia+GTX750 ) and maybe a new PSU. Questions: 1. With a card like this, how much acceleration (very approximately) can we expect in a standard simulation (protein+lipid+water, PME electrostatics)? In PME simulations you should see about 3x speedup, more if you are lucky and possibly a little less if you are not so lucky with the computational workload. What FF are you using? 2. Should we put any extra considerations into cooling? We expect week-long runs, and I am not sure these cards have this type of hardcore gaming in mind. :) I don't think so, but it depends how stuffed or dusty is that Dell workstation. These Maxwell generation (96x) GPUs run quite cool. Any additional comments are highly appreciated. Please keep in mind that this is a toy project to utilize an old workstation. There is a GPU fan speed issue with desktop NVIDIA cards, the fan rotation is limited to 60% and cards can overheat. You can force the fan speed higher using nvidia-settings (you'll need an X server running, though). For more details you can check slide 28 here: https://goo.gl/7DnSri Cheers, -- Szilárd Thank you, Alex -- Gromacs Users mailing list * Please search the archive at http
Re: [gmx-users] GPU acceleration
Hi Szilárd, Well, yes, it is an old CPU, but I am actually quite content with the ns/day performance it churns out on all four cores. I'm looking at http://manual.gromacs.org/documentation/5.1-beta1/user-guide/environment-variables.html and it is my understanding that GMXRC needs to be changed to include those flags, but no recompiling is required. Is that correct? Thanks, Alex On Fri, Jun 5, 2015 at 1:52 PM, Szilárd Páll pall.szil...@gmail.com wrote: Hi Alex, That's a quite old CPU so it's hard to tell how fast is it, you'll have to try. There are ways to guesstimate how fast does your GPU have to be to keep up with your CPU cores (hint search for GMX_EMULATE_GPU / GMX_NO_NONBONDED on the list). In any case, unless your CPU is so slow that it balances well with the 750Ti, I would strongly recommend that you get a GTX 960 which will cost just a bit more, but it will be 1.6-1.8x faster at computing non-bonded interactions. On Fri, Jun 5, 2015 at 9:08 PM, Alex nedoma...@gmail.com wrote: Hi all, We're about to start a bit of a toy project to test GPU-accelerated GMX. What I have here is my old workstation (Dell Precision T3500: Quad-core Xeon W3530 @ 2.8 GHz, 12 gigs of RAM) running Linux Mint 17.1. The plan is to get something like GeForce GTX 750 ( http://www.amazon.com/GeForce-750Ti-GDDR5-Graphics-GTX750TI-OC-2GD5/dp/B00IB9P1KG/ref=sr_1_3?ie=UTF8qid=1433529993sr=8-3keywords=Nvidia+GTX750 ) and maybe a new PSU. Questions: 1. With a card like this, how much acceleration (very approximately) can we expect in a standard simulation (protein+lipid+water, PME electrostatics)? In PME simulations you should see about 3x speedup, more if you are lucky and possibly a little less if you are not so lucky with the computational workload. What FF are you using? 2. Should we put any extra considerations into cooling? We expect week-long runs, and I am not sure these cards have this type of hardcore gaming in mind. :) I don't think so, but it depends how stuffed or dusty is that Dell workstation. These Maxwell generation (96x) GPUs run quite cool. Any additional comments are highly appreciated. Please keep in mind that this is a toy project to utilize an old workstation. There is a GPU fan speed issue with desktop NVIDIA cards, the fan rotation is limited to 60% and cards can overheat. You can force the fan speed higher using nvidia-settings (you'll need an X server running, though). For more details you can check slide 28 here: https://goo.gl/7DnSri Cheers, -- Szilárd Thank you, Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] GPU acceleration
In that case, I am a bit lost, as I was sure GMXRC needed modification. :)). Since I never ran anything on GPU and have an mdrun line that looks like mdrun -nt 4 -s tpr -c final structure what should be added to trigger the behavior you're suggesting? I understand that it will likely blow up the system, but for the heck of it... I am now looking at the stats and see that (NxN QSTab + LJ) and (NxN QSTab [F]) take up about 86% of all flops in my simulation. Which of those get offloaded? Thank you, Alex On Fri, Jun 5, 2015 at 2:26 PM, Szilárd Páll pall.szil...@gmail.com wrote: Hi Alex, Why do you think GMXRC needs to be changed? You simply need to set those env. vars to trigger the GPU code-path - and you will likely have to freeze the system, otherwise it will probably not run more than a few steps without seeing an explosion. Based on this you'll know how fast you can run with GPU offload, but you can also guesstimate how fast of a GPU you need (based on the PME+Bonded force calculation time + publicly available CUDA kernel performance). Cheers, -- Szilárd On Fri, Jun 5, 2015 at 10:13 PM, Alex nedoma...@gmail.com wrote: Hi Szilárd, Well, yes, it is an old CPU, but I am actually quite content with the ns/day performance it churns out on all four cores. I'm looking at http://manual.gromacs.org/documentation/5.1-beta1/user-guide/environment-variables.html and it is my understanding that GMXRC needs to be changed to include those flags, but no recompiling is required. Is that correct? Thanks, Alex On Fri, Jun 5, 2015 at 1:52 PM, Szilárd Páll pall.szil...@gmail.com wrote: Hi Alex, That's a quite old CPU so it's hard to tell how fast is it, you'll have to try. There are ways to guesstimate how fast does your GPU have to be to keep up with your CPU cores (hint search for GMX_EMULATE_GPU / GMX_NO_NONBONDED on the list). In any case, unless your CPU is so slow that it balances well with the 750Ti, I would strongly recommend that you get a GTX 960 which will cost just a bit more, but it will be 1.6-1.8x faster at computing non-bonded interactions. On Fri, Jun 5, 2015 at 9:08 PM, Alex nedoma...@gmail.com wrote: Hi all, We're about to start a bit of a toy project to test GPU-accelerated GMX. What I have here is my old workstation (Dell Precision T3500: Quad-core Xeon W3530 @ 2.8 GHz, 12 gigs of RAM) running Linux Mint 17.1. The plan is to get something like GeForce GTX 750 ( http://www.amazon.com/GeForce-750Ti-GDDR5-Graphics-GTX750TI-OC-2GD5/dp/B00IB9P1KG/ref=sr_1_3?ie=UTF8qid=1433529993sr=8-3keywords=Nvidia+GTX750 ) and maybe a new PSU. Questions: 1. With a card like this, how much acceleration (very approximately) can we expect in a standard simulation (protein+lipid+water, PME electrostatics)? In PME simulations you should see about 3x speedup, more if you are lucky and possibly a little less if you are not so lucky with the computational workload. What FF are you using? 2. Should we put any extra considerations into cooling? We expect week-long runs, and I am not sure these cards have this type of hardcore gaming in mind. :) I don't think so, but it depends how stuffed or dusty is that Dell workstation. These Maxwell generation (96x) GPUs run quite cool. Any additional comments are highly appreciated. Please keep in mind that this is a toy project to utilize an old workstation. There is a GPU fan speed issue with desktop NVIDIA cards, the fan rotation is limited to 60% and cards can overheat. You can force the fan speed higher using nvidia-settings (you'll need an X server running, though). For more details you can check slide 28 here: https://goo.gl/7DnSri Cheers, -- Szilárd Thank you, Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org
Re: [gmx-users] Microwave
I concur, this is a normal practice. Although I don't really get it, because GMX has a built-in capability of applying AC electric fields. Which by the way opens great doors to tinfoil hat research! Will someone please show that cellphone signal intensity is enough to make ion channels go nuts? A very necessary field of research, IMO. Alex MA Hi, MA You'd have to contact the authors to get their modified version. MA Mark MA On Sun, Jun 7, 2015 at 11:09 PM Mostafa Javaheri javaheri.grom...@gmail.com MA wrote: Hi all, Thanks for your replies, I searched about the simulation under e/m field and as far as I understood, I have to alter the source code for that, but I don't know how to do this and which part of the source code must change exactly. several studies have been performed by Niall J. English and co ( http://scitation.aip.org/content/aip/journal/jcp/131/3/10.1063/1.3184794) and (http://adsabs.harvard.edu/abs/2003JChPh.118.1589E) which I'm tending to do the same but with different protein, due to the correspondence, they used modified version of GROMACS 3 in their studies. I would be thankful if you could give me some idea about this kind of modification? Best, Mostafa On Tue, Jun 2, 2015 at 4:37 PM, Justin Lemkul jalem...@vt.edu wrote: On 6/2/15 4:31 AM, Mostafa Javaheri wrote: Dear Justin, How could I simulate a membrane protein while there is microwave radiations in one direction e.g. in direction of z-axis? Should I change the delta-lambda delta-lambda is a free energy option. or Is this accessible via electric field option? If I have to use electric field option, should I patch the mdrun ( http://www.gromacs.org/Developer_Zone/Programming_Guide/Patching_mdrun )? Electric fields are already supported, but I don't think this is physically what you want. I have no idea how to microwave an MD simulation. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] About Pulling code in Gromacs 5.0.2
The artifacts warning is general (mentioned in the manual) and has nothing to do with your system (aside from your mdp), everyone using absolute reference frame gets this warning. That said, the error you're getting is actually descriptive. So: 1. Check if the system is stable and happy without pulling. 2. Make sure the pulling rate is compatible with your nstlist setting. If the neighbor list update rate can't keep up with your pulling rate, that's a problem. High pulling rates in thermostatted systems are also generally problematic from the physical standpoint. Severely problematic. Steered MD is to be carefully designed. 3. Make sure your boundaries are taken care of properly, in case your pulled object meets the wall. Alex hy Hello all, hy Recently, I preformed a SMD, pulling a nano diamond across the membrane. I hy found that pull_geometry position in the Gromacs 4.6.5 (suggested in the hy tips of the US tutorial by Justin.A ) was replaced by pulling with absolute hy reference point in Gromacs 5.0.2. After defined pulling parameters hy according to the manual, Gromacs gave me a warning that This will lead to hy artifacts. I really want to know what did it mean, so I ignored it by hy using maxwarn 1 and saw what would happen. Finally I got a segment fault hy that a bonded atom moved away, out of the cut-off distance. hy Can you give me some advises? How can I treat the pulling correctly for hy this penetrating process and why does the absolute reference lead to hy artifacts. hy Many thanks! hy Hang -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] GPU acceleration
Thanks for your help, but nevermind this whole thing. I incidentally found out that our cluster had a node with two Tesla cards, which I am now using with GMX 5.0.5 freshly compiled with the NVIDIA library. I am getting massive (up to x10) acceleration compared with CPU-only runs. I think I posted yesterday about a fatal error due to CPU-GPU sync issue, but seems like I was able to take care of it. Very impressed here, about to abandon all of my CPU-only runs and switch to the GPU node as my workhorse. Alex On Tue, Jun 9, 2015 at 3:43 PM, Szilárd Páll pall.szil...@gmail.com wrote: No, it's not. Is this 5.0 or master branch? -- Szilárd On Fri, Jun 5, 2015 at 11:48 PM, Alex nedoma...@gmail.com wrote: Thanks for the info. I tried with GMX_EMULATE_GPU=1 preceding the mdrun line, crashed after 100 steps. Is that the expected behavior? On Fri, Jun 5, 2015 at 2:46 PM, Szilárd Páll pall.szil...@gmail.com wrote: On Fri, Jun 5, 2015 at 10:39 PM, Alex nedoma...@gmail.com wrote: In that case, I am a bit lost, as I was sure GMXRC needed modification. :)). Since I never ran anything on GPU and have an mdrun line that looks like mdrun -nt 4 -s tpr -c final structure gmx mdrun -ntomp 4 No need to do domain decomposition. Even with CPU runs it'll be faster if you don't. what should be added to trigger the behavior you're suggesting? I understand that it will likely blow up the system, but for the heck of it... those environment variables I mentioned. Plus timer resetting, e.g. with -resethway if you want reliable timings. I am now looking at the stats and see that (NxN QSTab + LJ) and (NxN QSTab [F]) take up about 86% of all flops in my simulation. Which of those get offloaded? All of that will be offloaded. However, those are theoretical (useful) flops, the timing table on the bottom is more relevant. Set the GMX_CYCLE_SUBCOUNTER=ON breakdown of the Force. However, if you just follow the previous emulate+skip non-bondeds advise, you'll get a run that does the offload except that it skips calculating - which is useful because you have no GPU to calculate with. Thank you, Alex On Fri, Jun 5, 2015 at 2:26 PM, Szilárd Páll pall.szil...@gmail.com wrote: Hi Alex, Why do you think GMXRC needs to be changed? You simply need to set those env. vars to trigger the GPU code-path - and you will likely have to freeze the system, otherwise it will probably not run more than a few steps without seeing an explosion. Based on this you'll know how fast you can run with GPU offload, but you can also guesstimate how fast of a GPU you need (based on the PME+Bonded force calculation time + publicly available CUDA kernel performance). Cheers, -- Szilárd On Fri, Jun 5, 2015 at 10:13 PM, Alex nedoma...@gmail.com wrote: Hi Szilárd, Well, yes, it is an old CPU, but I am actually quite content with the ns/day performance it churns out on all four cores. I'm looking at http://manual.gromacs.org/documentation/5.1-beta1/user-guide/environment-variables.html and it is my understanding that GMXRC needs to be changed to include those flags, but no recompiling is required. Is that correct? Thanks, Alex On Fri, Jun 5, 2015 at 1:52 PM, Szilárd Páll pall.szil...@gmail.com wrote: Hi Alex, That's a quite old CPU so it's hard to tell how fast is it, you'll have to try. There are ways to guesstimate how fast does your GPU have to be to keep up with your CPU cores (hint search for GMX_EMULATE_GPU / GMX_NO_NONBONDED on the list). In any case, unless your CPU is so slow that it balances well with the 750Ti, I would strongly recommend that you get a GTX 960 which will cost just a bit more, but it will be 1.6-1.8x faster at computing non-bonded interactions. On Fri, Jun 5, 2015 at 9:08 PM, Alex nedoma...@gmail.com wrote: Hi all, We're about to start a bit of a toy project to test GPU-accelerated GMX. What I have here is my old workstation (Dell Precision T3500: Quad-core Xeon W3530 @ 2.8 GHz, 12 gigs of RAM) running Linux Mint 17.1. The plan is to get something like GeForce GTX 750 ( http://www.amazon.com/GeForce-750Ti-GDDR5-Graphics-GTX750TI-OC-2GD5/dp/B00IB9P1KG/ref=sr_1_3?ie=UTF8qid=1433529993sr=8-3keywords=Nvidia+GTX750 ) and maybe a new PSU. Questions: 1. With a card like this, how much acceleration (very approximately) can we expect in a standard simulation (protein+lipid+water, PME electrostatics)? In PME simulations you should see about 3x speedup, more if you are lucky and possibly a little less if you are not so lucky with the computational workload. What FF are you using? 2. Should we put any extra considerations
Re: [gmx-users] GPU acceleration
At this point, we will be getting the card for sure, likely the GTX 960 you suggested. Hence, a question. Which of the manufacturers would you recommend in terms of cooling (if you think it might become an issue with week or so long runs)? I doubt we'll be needing the 4GB version, so my only concern is cooling. Comments highly appreciated. Thanks, Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] constraint pull code / gromacs 5.0.4
CA My nanoparticles (Radius = 1nm) are made of alumina and CA the force field was CLAYFF I assume you realize that simulating metals within the GMX-like paradigm is physically incorrect. In short, your particles are not metallic, but crystalline clusters with vdw surface. That in itself, however, should not cause what you're observing. Let's try to diagnose it one step at a time. Can you please look at the pdb and post the first-neighbor interatomic AL-AL distance? Also, please post the bond length and the equilibrium angle in the forcefield parameters. Is it possible to simulate one of these nanoparticles in vacuum (no pull code, no solvent, NVE) and look at the kinetic/potential energies of the system? Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Get pdb2gmx not to treat the first nucleotide in a chain as 5'
Peter, it should be in aminoacids, just a legacy file naming. Alex On Mon, Jun 22, 2015 at 12:46 PM, Peter Stern peter.st...@weizmann.ac.il wrote: Even though it's called aminoacids.n.tdb? Or a different name, e.g. nucleicacids.n.tdb? One problem is that I eventually have to deal with tens, maybe hundreds of complexes. Some, of course, will have the normal 5' nucleotide. Will I be able to choose the appropriate n.tdb file for each case or will I have to copy the appropriate one to my run directory each time? I guess the latter. Thanks, Peter -Original Message- From: gromacs.org_gmx-users-boun...@maillist.sys.kth.se [mailto: gromacs.org_gmx-users-boun...@maillist.sys.kth.se] On Behalf Of Justin Lemkul Sent: Monday, June 22, 2015 9:39 PM To: gmx-us...@gromacs.org Subject: Re: [gmx-users] Get pdb2gmx not to treat the first nucleotide in a chain as 5' On 6/22/15 2:36 PM, Peter Stern wrote: I have a chain of RNA in complex with a protein. The pdb file is missing the first two nucleotides and starts with the third including the PO2 group. pdb2gmx identifies this as the 5' terminal nucleotide and tries to create the topology from rna.rtp using RA5 as the residue type. Thus it doesn't like the fact that the coordinates contain a P atom and fails. I would like to avoid removing the coordinates for PO2. How can I get pdb2gmx to treat the first (and last in other cases) nucleotide as a normal residue? Simply changing the name of the residue from A to RA does not work. You'll need to create an entry in the .n.tdb file that specifies a 5'-phosphate cap. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] GPU acceleration
Well, I've had a Gigabyte-made OC'd GTX 660 at home (for gaming) for a few years, no problems at all. I guess my question was more along the lines of whether GPU MD runs are equivalent to week-long gaming in terms of heat production. What I also didn't know is whether 960 was a fiery beast. Your post answers that and we'll be getting one, probably from ASUS. I try to stay away from EVGA (terrible experience, card was returned, but I kept the Powered by EVGA sticker, which is now proudly presented on my toilet). Thanks. Alex SP I can't really give you much recommendation in terms of manufacturer. SP I typically suggest people to stay away from factory overclocked cards SP (although some have been reporting stable use for years of such SP cards). Otherwise the GTX 960 is not a beast in terms of power SP consumption and I would not expect cooling to be much of an issue. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] GMX 5.0.5 x2top segfaults
Mark, I can't really share that particular setup, unfortunately. Let me see if it fails on something I can share. Alex Hi, Having a set of input files that reproduce the problem with x2top uploaded to that redmine will make it more likely that someone will attempt a fix. I'm not aware that any of the core devs uses x2top, so without the ability to reproduce the problem and show it is fixed, people will probably find something else to do... Mark On Tue, Jun 16, 2015 at 1:34 PM Justin Lemkul jalem...@vt.edu wrote: On 6/15/15 10:21 PM, Alex wrote: Ahoy all. What's going on with the latest version of x2top? It spectacularly segfaults at the point of doing dihedrals. Identical setup (local oplsaa.ff with some customization) works flawlessly with 4.6.5 and the resulting topologies are all perfectly reasonable. Not cool. Any suggestions, besides having an old version around? I already have that, and it's a bit less than kosher. Known bug. http://redmine.gromacs.org/issues/1711 Until someone has time to fix it, you have to use something in the 4.6.x series. In trying to fix pdb2gmx to do funny things like having angles without bonds, that broke x2top. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu| (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive athttp://www.gromacs.org/Support/Mailing_Lists/GMX-Users_Listbefore posting! * Can't post? Readhttp://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-usersor send a mail togmx-users-requ...@gromacs.org. -- Best regards, Alex mailto:nedoma...@gmail.com -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
[gmx-users] GMX 5.0.5 x2top segfaults
Ahoy all. What's going on with the latest version of x2top? It spectacularly segfaults at the point of doing dihedrals. Identical setup (local oplsaa.ff with some customization) works flawlessly with 4.6.5 and the resulting topologies are all perfectly reasonable. Not cool. Any suggestions, besides having an old version around? I already have that, and it's a bit less than kosher. Thanks, Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Microwave
It really depends on what level of theory you want to have in your simulation. If this is simple temperature increase due to GHz AC electric fields, GMX has the provisions already, and there's no need to patch anything. If you need magnetic fields and explicit secondary radiation, i.e. MD coupled with the FDTD method for solving charge-coupled Maxwell's equations, I am not aware of any tools that can do this. Which is kind of strange, because it is both doable and useful. Alex MJ Dear Justin, MJ How could I simulate a membrane protein while there is microwave radiations in MJ one direction e.g. in direction of z-axis? Should I change the delta-lambda MJ or Is this accessible via electric field option? If I have to use electric MJ field option, should I patch the mdrun ( MJ http://www.gromacs.org/Developer_Zone/Programming_Guide/Patching_mdrun)? MJ I will appreciate it if you could answer these questions. MJ Best, MJ Mostafa -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] GPU and aux power supply
I am happy to say that I am getting an 8-fold increase in simulation speeds for $200. An additional question: normally, how many simulations (separate mdruns on separate CPU cores) can be performed simultaneously on a single GPU? Say, for 20-40K particle sized simulations. The coolers are not even spinning during a single test (mdrun -ntomp 4), and I get massive acceleration. They aren't broken, the card is just cool (small system, ~3K particles). Thanks, Alex Ah, ok, so you can get a 6-pin from the PSU and another from a converted molex connector. That should be just fine, especially as the card should will not pull more than ~155W (under heavy graphics load) based on the Tomshardware review* and you are providing 225W max. *http://www.tomshardware.com/reviews/evga-super-super-clocked-gtx-960,4063-3.html -- Szilárd On Tue, Jun 30, 2015 at 7:31 PM, Alex nedoma...@gmail.com wrote: Well, I don't have one like this. What I have instead is this: 1. A single 6-pin directly from the PSU. 2. A single molex to 6-pin (my PSU does provide one molex). 3. Two 6-pins to a single 8-pin converter going to the card. In other words, I can populate both 6-pins on the 6-8 converter, just not sure about the pinouts in this case. Not good? Alex What I meant is this:http://goo.gl/8o1B5P That is 2x molex - 8pin PCI-E. A single molex may not be enouhg. -- Szilárd On Tue, Jun 30, 2015 at 7:10 PM, Alex nedoma...@gmail.com wrote: It is a 4-core CPU, single GPU box, so I doubt I will be running more than one at a time. We will very likely get a different PSU, unless... I do have a molex to 6 pin concerter sitting on this very desk. Do you think it will satisfy the card? I just don't know how much a single molex line delivers. If you feel this should work, off to installing everything I go. Thanks a bunch, Alex SP First of all, unless you run multiple independent simulations on the same SP GPU, GROMACS runs alone will never get anywhere near the peak power SP consumption of the GPU. SP The good news is that NVIDIA has gained some sanity and stopped blocking SP GeForce GPU info in nvidia-smi - although only for newer cars, but it does SP work with the 960 if you use a 352.xx driver: SP +--+ SP | NVIDIA-SMI 352.21 Driver Version: 352.21 | SP |---+--+--+ SP | GPU NamePersistence-M| Bus-IdDisp.A | Volatile Uncorr. SP ECC | SP | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute SP M. | SP |===+==+==| SP | 0 GeForce GTX 960 Off | :01:00.0 On | SP N/A | SP | 8% 45C P5 15W / 130W | 1168MiB / 2044MiB | 31% SP Default | SP +---+--+--+ SP A single 6-pin can deliver 75W, an 8-pin 150W, so in your case, the hard SP limits of what your card can pull is 75W from the PCI-E slow + 150W from SP the cable = 225 W. With a single 6-pin cable you'll only get ~150W max. SP That can be OK if your card does not pull more power (e.g. the above SP non-overclocked card would be just fine), but as your card is overclocked, SP I'm not sure it won't peak above 150W. SP You can try to get a molex - PCI-E power cable converter. SP -- SP Szilárd SP On Mon, Jun 29, 2015 at 9:56 PM, Alex nedoma...@gmail.com wrote: Hi all, I have a bit of a gromacs-unrelated question here, but I think this is a better place to ask it than, say, a gaming forum. The Nvidia GTX 960 card we got here came with an 8-pin AUX connector on the card side, which interfaces _two_ 6-pin connectors to the PSU. It is a factory superclocked card. My 525W PSU can only populate _one_ of those 6-pin connectors. The EVGA website states that I need at least 400W PSU, while I have 525. At the same time, I have a dedicated high-power PCI-e slot, which on the motherboard says "75W PCI-e". Do I need a different PSU to populate the AUX power connector completely? Are these runs equivalent to drawing max power during gaming? Thanks! Alex -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_Listbefore posting! * Can't post? Readhttp://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-usersor send a mail togmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive athttp://www.gromacs.org/Support/Mailing_Lists/GMX-Users_Listbefore posting! * Can't post? Readhttp://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-usersor send a mail togmx-users-requ...@gromacs.org. -- Best regards, Alex mailto:nedoma...@gmail.com -- Best reg
