Hi, 15 apr 2010 kl. 11.11 skrev Ramachandran G:
> Hello Par: > Using the latest git i could able to work on my oxy-hemoglobin system > with new gromacs version 4.0.5 successfully. CHARMM is not supportd in version 4.0.5, you probably mean the developer version? > But since my new gromacs version was not installed in parallel i tried using > older version of gromacs 3.3.1 which was installed in parrallel. While doing > grompp to get the *.tpr file, i am getting the below error. > ------------------------------------------------------------------------------------------------- > Program grompp, VERSION 3.3.1 > Source code file: topdirs.c, line: 103 > > Fatal error: > Invalid dihedral type 9 > -------------------------------------------------------------------------------------------------- This is because we have made changes in the source code to allow for the CHARMM ff (we added a new dihedral type, see paper, that is causing this particular error). So, you have to run with the developer git version. Regards, Pär > > I am planning to install the newer gromacs version also in parallel but in > the mean time i also wanted to rectify the above error, if i can. Thank you. > > Rama > > > On Thu, Mar 18, 2010 at 3:02 AM, Pär Bjelkmar <[email protected]> wrote: > Hi, > > I have now added the charmm files to git head and the protein and lipid parts > should work at least. Simply check out the latest git and the charmm > parameters should show up in pdb2gmx. > > /Pär > > 11 mar 2010 kl. 03.04 skrev Ramachandran G: > >> Hello Bjelkumar, >> Thanks for your reply. I will wait until you add the new force >> field to the git. >> >> regards, >> rama >> >> On Tue, Mar 2, 2010 at 12:47 AM, Pär Bjelkmar <[email protected]> wrote: >> Hello, >> >> I have that on my schedule for next week. That is to add the new force field >> format to the git source code tree. Until then you could use the old format >> if you checkout a previous version of the source code. I guess you know >> about git otherwise there's some info here >> http://www.gromacs.org/Developer_Zone/Git/Basic_Git_Usage >> >> So start by getting a copy of the latest _developing_ version (git head) of >> gromacs source code: >> $ git clone [email protected]:gromacs.git >> The force field files were update in end of January to you could checkout a >> version from Jan 27 by: >> $ git checkout d80138d720fd41a5005238adecd5bfd5400e524c >> >> then compile. Now it should work if you got the HEME parameters from Michel. >> Alternatively you wait until end of next week when I think I have had time >> to add the new force field format to the git head. >> >> Good luck! >> /Pär >> >> 1 mar 2010 kl. 21.21 skrev Ramachandran G: >> >>> Dear Pär Bjelkmar, >>> After getting help from Michel Cuendet still i am facing problem >>> and i >>> understood that in groamcs latest version 4.0.x problems may come due to >>> CMAP. >>> He suggested me to seek your help. Could you help me to get new toplogy >>> format charmm files which can adapt the HEME sections to that. >>> Thank you. >>> >>> Rama >>> >>> ---------- Forwarded message ---------- >>> From: Michel Cuendet <[email protected]> >>> Date: Fri, Feb 26, 2010 at 1:03 PM >>> Subject: Re: ffcharmm27 for HEME >>> To: Ramachandran G <[email protected]> >>> >>> >>> >>> Dear Mr. Ramachandran, >>> >>> This means you are using grompp from the distribution version 4.0.x. >>> This version of gromacs does not know about the CMAP forcefield term. >>> You have two options : >>> >>> 1) Deactivate CMAP. This would be equivalent to using the charmm22 ff >>> instead of the charmm27, which many people consider fine. To do this, >>> edit the file ffcharmm27.rtp and delete all [cmap] sections. This >>> _should_ work. >>> >>> 2) Download the git development version, which knows how to handle CMAP. >>> I know they have been changing topology formats in the latest versions, >>> but I hope it is still able to read the old topologies. Otherwise, you >>> will have to ask Par Bjelkmar for the charmm files in the new topology >>> format, and adapt the HEME sections to that. >>> >>> Sorry it is a bit messy at the moment. But everything should become part >>> of the gormacs4.1 distribution. >>> >>> Bye, >>> Michel >>> >>> Ramachandran G wrote: >>> > Dear Mr. Michel, >>> > Thank you very much for you help and time. >>> > Although i could able to do pdb2gmx and get the *.top file still, i >>> > have problem in getting the 'tpr' file using grompp. This is due the >>> > 'cmap' >>> > >>> > I would try to fix the problem myself but still if the problem persist >>> > then i will get back to you. Thank you again. >>> > >>> > regards, >>> > Rama >>> > >>> > On Thu, Feb 25, 2010 at 8:59 AM, Michel Cuendet >>> > <[email protected] <mailto:[email protected]>> wrote: >>> > >>> > >>> > Dear Mr. Ramachandran, >>> > >>> > Attached is a tar file containing the ffcharmm27 forcefield files >>> > modified for the HEME and an optional CO ligand. >>> > >>> > Note that I also wanted to add the O2 ligand, but the original charmm >>> > parameters seem to be inconsistent. For example, the FE-OM-OM angle is >>> > 180deg, which is wrong. And there is no angle defined for NH2-FE-OM, >>> > while a strong angle is required to keep the ligands axial. So we have >>> > corrected this and guessed a couple of missing parameters for our own >>> > use. But I do not include them to the official ffcharmm27 >>> > distribution, >>> > which should stay as close as possible to the original charmm >>> > force field. >>> > >>> > There is a subtlety with the histidine binding to HEME. The Charmm >>> > patch >>> > PHEM defines a dihedral term only with NA and not with NB,NC, or ND. >>> > ; 1CD2 1NE2 2FE 2NA (1 = HSD, 2 = HEME) >>> > In gromacs we cannot differentiate between NA, NB, NC, and ND of the >>> > heme (and special values cannot be attributed in the rtp file for >>> > dihedrals involving atoms from other residues). Given the symmetry of >>> > the HEME we can distribute the dihedral over NA, NB, NC, and ND. So we >>> > divide the original force constant (0.2092) by four. This should give >>> > pretty much the same behavior in practice. >>> > >>> > In order to treat the HEME with pdb2gmx, you have to (assuming the >>> > HEME >>> > is covalently bound to an histidine in chain A): >>> > >>> > 1) Attribute HEME to chain A (Covalent bonds only within one chain) >>> > >>> > 2) Move HEME to end of chain A (all atoms of one chain have to be >>> > consecutive for gromacs) >>> > >>> > See the attached pdb file as an example. >>> > >>> > I hope it will work. Do not hesitate to contact me if you have any >>> > additional questions, >>> > >>> > Best regards, >>> > Michel >>> > >>> > >>> > >>> > > *Från: *Ramachandran G <[email protected] >>> > <mailto:[email protected]> <mailto:[email protected] >>> > <mailto:[email protected]>>> >>> > > *Datum: *22 februari 2010 23.56.42 CET >>> > > *Till: *[email protected] <mailto:[email protected]> >>> > <mailto:[email protected] <mailto:[email protected]>> >>> > > *Ämne: **ffcharmm* >>> > > >>> > > Hi Pär Bjelkmar, >>> > > Thank you for the charmm force field patches for gromacs. It is >>> > > very much useful for me. However presently i am facing a problem >>> > since >>> > > these force fields do not have 'HEME' group in it. I tried to >>> > add it >>> > > manually but unsuccessful. Could you help me in adding the HEME >>> > group >>> > > to the ffcharmm patches. Your help would be highly appreciated. >>> > > >>> > > >>> > > Ramachandran. G >>> > > -- >>> > > Postdoctoral Research Scholar, >>> > > Department of Chemistry, >>> > > University of Nevada, Reno. >>> > >>> > >>> > >>> > >>> > -- >>> > Postdoctoral Research Scholar, >>> > Department of Chemistry, >>> > University of Nevada, Reno. >>> >>> >>> >>> >>> -- >>> Postdoctoral Research Scholar, >>> Department of Chemistry, >>> University of Nevada, Reno. >> >> >> >> >> -- >> Postdoctoral Research Scholar, >> Department of Chemistry, >> University of Nevada, Reno. > > > > >
-- gmx-users mailing list [email protected] http://lists.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to [email protected]. Can't post? Read http://www.gromacs.org/mailing_lists/users.php
