Re: [gmx-users] Velocities from the .gro file
Many thanks for your reply :) All your language assumptions are true and that is exactly what I wanted to communicate, next time I will try to be more precise and sorry for the confusion I will read section 3.4.1 again carefully. Thanks again and sorry for the inconvenience. Mohamed On Thu, Apr 9, 2020 at 04:33 Eric Smoll wrote: > On Wed, Apr 8, 2020 at 4:41 PM Mohamed Abdelaal > wrote: > > > Many thanks for your reply > > > > The limitation in the generate velocity using the .mdp file, is that > while > > I can generate the velocity from Maxwell distribution, I will have the > > same velocities in the x, y and z directions. > > > > I think you mean "same velocity *distributions* in the x, y, and z > directions." The distributions will be approximately the same but each > atom will have a different velocity. > > > > > On the other hand, generating the velocity from the .gro file will let me > > specify different velocities in the x,y and z directions but they will be > > the same velocities for all the atoms (will not be taken from a maxwell > > distribution with variation in the atoms velocities). > > > > I think you mean "specify different velocity *distributions* in the x, y, > and z directions" > > > > > Is it possible to generate different velocities in the x,y and z > directions > > > > I think you mean "generate different velocity *distributions* in the x, y, > and z directions." If so, the answer is obviously yes. Because you can > type in each individual vxi, vyi, and vzi for every atom i, you can > generate different velocity distributions in the x, y, and z directions. > > > > from a maxwell distribution ? > > > I am not sure what this part of the sentence means. If you do what you are > suggesting, you will not be working with a maxwell distribution because all > three directions should have identical distributions. See comment below. > If there is another misunderstanding, you need to spend more time crafting > precise sentences to communicate what you are after. > > > > (for example the velocities to be taken from > > a maxwell distribution with a mean of 0.1 in the x direction and with a > > mean of 0.2 in the y direction and with mean of 0.3 in the z direction?) > > > > In my last email I suggested reading section 3.4.1 of the manual version > 5.1.2. It seems you did not. A 3D Maxwell Boltzmann velocity distribution > corresponds to three identical gaussian speed distributions in vx, vy, and > vz centered at zero (mean should be zero for vx, vy, vz). Just change the > standard deviation of the velocity distribution sqrt(kT/m) for each > velocity component if you want them to be different. If you don't want the > mean to be zero for whatever reason, add a constant. However, a non-zero > mean for any of the velocity components will generate center of mass > motion. If you want center of mass motion, turn off center of mass motion > removal in the mdp file. > > > > Thanks for your help :) > > Mohamed > > > > On Wed, Apr 8, 2020 at 05:48 Eric Smoll wrote: > > > > > On Tue, Apr 7, 2020 at 3:38 PM Mohamed Abdelaal < > m.b.abdel...@gmail.com> > > > wrote: > > > > > > > No, I use the generate velocity option in the .mdp files. > > > > > > > > However I want now to assign different velocities in the x,y,z > > > directions. > > > > Which I thought it could only be done through the .gro file, but I > > don't > > > > know If I did that, should I change the value of the generate > velocity > > to > > > > be = NO in the .mdp files ? (otherwise I would have generated the > > > > velocities twice). > > > > > > > > > > That sounds logical. Set it to no if you provide your own initial > > > velocities. > > > > > > > > > > > Moreover, if I added the velocities in the .gro file how can I > generate > > > the > > > > velocities in the .gro file from a distribution (Maxwell) with a > > specific > > > > mean and standard deviation ? > > > > > > > > I have tried to search in different sources (the user list, manual, > > user > > > > guide, research gate and other platforms) how to solve this velocity > > > > problem but I didn't find a clear way to insert different velocities > in > > > the > > > > x,y,z directions using distribution rater than constant velocities. > > > > > > > > There is a good section on this in the manual. For example, section > > > 3.4.1 > > > in the Gromacs 5.1.2 manual. > > > > > > Also, you know that the generate velocities option assigns velocities > to > > > atoms from an approximate MB distribution at whatever temperature you > > > specify in the MDP file, right? If I understand you correctly, the > > > generate velocities options should do exactly what you want. With no > > extra > > > work. > > > > > > > > > > Please guide me how to do it as I am a little bit confused in the > > > velocity > > > > generation mechanisms. > > > > > > > > Many thanks, > > > > Mohamed > > > > > > > > On Mon, Apr 6, 2020 at 6:19 PM Justin Lemkul > wrote: > > > > > > > > > > >
Re: [gmx-users] Velocities from the .gro file
On Wed, Apr 8, 2020 at 4:41 PM Mohamed Abdelaal wrote: > Many thanks for your reply > > The limitation in the generate velocity using the .mdp file, is that while > I can generate the velocity from Maxwell distribution, I will have the > same velocities in the x, y and z directions. > I think you mean "same velocity *distributions* in the x, y, and z directions." The distributions will be approximately the same but each atom will have a different velocity. > > On the other hand, generating the velocity from the .gro file will let me > specify different velocities in the x,y and z directions but they will be > the same velocities for all the atoms (will not be taken from a maxwell > distribution with variation in the atoms velocities). > I think you mean "specify different velocity *distributions* in the x, y, and z directions" > > Is it possible to generate different velocities in the x,y and z directions > I think you mean "generate different velocity *distributions* in the x, y, and z directions." If so, the answer is obviously yes. Because you can type in each individual vxi, vyi, and vzi for every atom i, you can generate different velocity distributions in the x, y, and z directions. > from a maxwell distribution ? I am not sure what this part of the sentence means. If you do what you are suggesting, you will not be working with a maxwell distribution because all three directions should have identical distributions. See comment below. If there is another misunderstanding, you need to spend more time crafting precise sentences to communicate what you are after. > (for example the velocities to be taken from > a maxwell distribution with a mean of 0.1 in the x direction and with a > mean of 0.2 in the y direction and with mean of 0.3 in the z direction?) > In my last email I suggested reading section 3.4.1 of the manual version 5.1.2. It seems you did not. A 3D Maxwell Boltzmann velocity distribution corresponds to three identical gaussian speed distributions in vx, vy, and vz centered at zero (mean should be zero for vx, vy, vz). Just change the standard deviation of the velocity distribution sqrt(kT/m) for each velocity component if you want them to be different. If you don't want the mean to be zero for whatever reason, add a constant. However, a non-zero mean for any of the velocity components will generate center of mass motion. If you want center of mass motion, turn off center of mass motion removal in the mdp file. > Thanks for your help :) > Mohamed > > On Wed, Apr 8, 2020 at 05:48 Eric Smoll wrote: > > > On Tue, Apr 7, 2020 at 3:38 PM Mohamed Abdelaal > > wrote: > > > > > No, I use the generate velocity option in the .mdp files. > > > > > > However I want now to assign different velocities in the x,y,z > > directions. > > > Which I thought it could only be done through the .gro file, but I > don't > > > know If I did that, should I change the value of the generate velocity > to > > > be = NO in the .mdp files ? (otherwise I would have generated the > > > velocities twice). > > > > > > > That sounds logical. Set it to no if you provide your own initial > > velocities. > > > > > > > > Moreover, if I added the velocities in the .gro file how can I generate > > the > > > velocities in the .gro file from a distribution (Maxwell) with a > specific > > > mean and standard deviation ? > > > > > > I have tried to search in different sources (the user list, manual, > user > > > guide, research gate and other platforms) how to solve this velocity > > > problem but I didn't find a clear way to insert different velocities in > > the > > > x,y,z directions using distribution rater than constant velocities. > > > > > > There is a good section on this in the manual. For example, section > > 3.4.1 > > in the Gromacs 5.1.2 manual. > > > > Also, you know that the generate velocities option assigns velocities to > > atoms from an approximate MB distribution at whatever temperature you > > specify in the MDP file, right? If I understand you correctly, the > > generate velocities options should do exactly what you want. With no > extra > > work. > > > > > > > Please guide me how to do it as I am a little bit confused in the > > velocity > > > generation mechanisms. > > > > > > Many thanks, > > > Mohamed > > > > > > On Mon, Apr 6, 2020 at 6:19 PM Justin Lemkul wrote: > > > > > > > > > > > > > > > On 4/6/20 12:16 PM, Mohamed Abdelaal wrote: > > > > > Hello everybody :) > > > > > > > > > > Can I use the gmx insert-molecules to insert molecules in my box > with > > > > > velocities by adding the velocities in the .gro file and insert the > > > > > molecules from this .gro file ? > > > > > > > > Have you tried it? > > > > > > > > -Justin > > > > > > > > -- > > > > == > > > > > > > > Justin A. Lemkul, Ph.D. > > > > Assistant Professor > > > > Office: 301 Fralin Hall > > > > Lab: 303 Engel Hall > > > > > > > > Virginia Tech Department of
Re: [gmx-users] Velocities from the .gro file
Many thanks for your reply The limitation in the generate velocity using the .mdp file, is that while I can generate the velocity from Maxwell distribution, I will have the same velocities in the x, y and z directions. On the other hand, generating the velocity from the .gro file will let me specify different velocities in the x,y and z directions but they will be the same velocities for all the atoms (will not be taken from a maxwell distribution with variation in the atoms velocities). Is it possible to generate different velocities in the x,y and z directions from a maxwell distribution ? (for example the velocities to be taken from a maxwell distribution with a mean of 0.1 in the x direction and with a mean of 0.2 in the y direction and with mean of 0.3 in the z direction?) Thanks for your help :) Mohamed On Wed, Apr 8, 2020 at 05:48 Eric Smoll wrote: > On Tue, Apr 7, 2020 at 3:38 PM Mohamed Abdelaal > wrote: > > > No, I use the generate velocity option in the .mdp files. > > > > However I want now to assign different velocities in the x,y,z > directions. > > Which I thought it could only be done through the .gro file, but I don't > > know If I did that, should I change the value of the generate velocity to > > be = NO in the .mdp files ? (otherwise I would have generated the > > velocities twice). > > > > That sounds logical. Set it to no if you provide your own initial > velocities. > > > > > Moreover, if I added the velocities in the .gro file how can I generate > the > > velocities in the .gro file from a distribution (Maxwell) with a specific > > mean and standard deviation ? > > > > I have tried to search in different sources (the user list, manual, user > > guide, research gate and other platforms) how to solve this velocity > > problem but I didn't find a clear way to insert different velocities in > the > > x,y,z directions using distribution rater than constant velocities. > > > > There is a good section on this in the manual. For example, section > 3.4.1 > in the Gromacs 5.1.2 manual. > > Also, you know that the generate velocities option assigns velocities to > atoms from an approximate MB distribution at whatever temperature you > specify in the MDP file, right? If I understand you correctly, the > generate velocities options should do exactly what you want. With no extra > work. > > > > Please guide me how to do it as I am a little bit confused in the > velocity > > generation mechanisms. > > > > Many thanks, > > Mohamed > > > > On Mon, Apr 6, 2020 at 6:19 PM Justin Lemkul wrote: > > > > > > > > > > > On 4/6/20 12:16 PM, Mohamed Abdelaal wrote: > > > > Hello everybody :) > > > > > > > > Can I use the gmx insert-molecules to insert molecules in my box with > > > > velocities by adding the velocities in the .gro file and insert the > > > > molecules from this .gro file ? > > > > > > Have you tried it? > > > > > > -Justin > > > > > > -- > > > == > > > > > > Justin A. Lemkul, Ph.D. > > > Assistant Professor > > > Office: 301 Fralin Hall > > > Lab: 303 Engel Hall > > > > > > Virginia Tech Department of Biochemistry > > > 340 West Campus Dr. > > > Blacksburg, VA 24061 > > > > > > jalem...@vt.edu | (540) 231-3129 > > > http://www.thelemkullab.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. > > > > > -- > > 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] Double-well potential for bonded interaction
Hi Jacek, Gromacs allows for the implementation of any kind of bonded interaction through user-defined tabulated functions. You'll have to build a table for it, but it's definitely more straightforward than messing around with restraints. Check out section 4.2.14 of the manual: "Tabulated bonded interaction functions". Best, John > Hi, > > Does anybody know how to define for a specific bond a double-well > potential? > > I've looking into the definitions in the gromacs manual > (manual.gromacs.org/current/reference-manual/topologies/topology-file-formats.html > and > manual.gromacs.org/current/reference-manual/functions/bonded-interactions.html), > but for now I can only imagine using bond restraints in a similar manner > shown here for position restraints: > manual.gromacs.org/current/reference-manual/functions/restraints.html > > It feels like there should be a more elegant way of doing that. > > Thanks for your suggestions in advance! > > Best, > Jacek > > > Dr. Jacek Kozuch > > Freie Universität Berlin > Fachbereich Physik > Arnimallee 14 > Raum 1.1.35 > 14 195 Berlin > > > -- > 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. John Whittaker Ph.D. Candidate Department of Mathematics and Computer Science Freie Universität Berlin +49 0160 936 04221 -- 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] Double-well potential for bonded interaction
Hi, Does anybody know how to define for a specific bond a double-well potential? I've looking into the definitions in the gromacs manual (manual.gromacs.org/current/reference-manual/topologies/topology-file-formats.html and manual.gromacs.org/current/reference-manual/functions/bonded-interactions.html), but for now I can only imagine using bond restraints in a similar manner shown here for position restraints: manual.gromacs.org/current/reference-manual/functions/restraints.html It feels like there should be a more elegant way of doing that. Thanks for your suggestions in advance! Best, Jacek Dr. Jacek Kozuch Freie Universität Berlin Fachbereich Physik Arnimallee 14 Raum 1.1.35 14 195 Berlin -- 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] Simulate only one unit of the virus capsid while fixing its surrounding units
Dear Cheng, The paper you mentioned (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5524983/) uses NAMD (http://www.ks.uiuc.edu/Research/namd/) has somewhat different scalability properties than GROMACS. Regards, Benson On Wed, Apr 8, 2020, at 7:14 AM, ZHANG Cheng wrote: > Dear Andre, > > > Thank you. We are trying to use an adenovirus as a vaccine. As it is > not stable, we want to simulate it to identify the unstable regions > (e.g. flexible), so as to either engineering (e.g. mutation) it, or > adding excipients. > > > Simulating only one protein of the capsid is of course doable. But do > you think simulating one protein without its neighbours could reflect > its dynamics? Would its boundary residues behave very differently > compared to with neighbours? > --Original-- > From:"ZHANG Cheng"<272699...@qq.com; > Date:Wed, Apr 8, 2020 11:02 AM > To:"gromacs.org_gmx-users" > Subject:Re:Simulate only one unit of the virus capsid while > fixing its surrounding units > > > > Dear Justin and Andre, > > > Thank you for the advice. So can I ask how commonly the very large > virus capsid is simulated? A recent paper "Physical properties of the > HIV-1 capsid from all-atom molecular dynamics simulations" is using > 3880 GPU accelerated Cray-XK nodes, which is impossible for our > university to provide. > > > > > -- Original -- > From:"ZHANG Cheng"<272699...@qq.com; > Date:Tue, Apr 7, 2020 10:10 PM > To:"ZHANG > Cheng"<272699...@qq.com;"gromacs.org_gmx-users" > Subject:Re:Simulate only one unit of the virus capsid while > fixing its surrounding units > > > > Dear Andre, Thank you for the advice. Can I ask, > > > 1) Could you please clarify the concepts? I know "constraint" and > "restraint" are two different things in gromacs. And "fix" is another > term? How about "freezegrps"? > > > 2) It is okay that the computational time is not reduced, as now only > several proteins are simulated. If I simulate all the several protein > without any fixing, I worry they will lose their conformation. So > fixing the neighbours and only focusing on the protein in the centre > could be the solution. > > > > > > -- Original -- > From:"ZHANG Cheng"<272699...@qq.com; > Date:Tue, Apr 7, 2020 09:41 PM > To:"gromacs.org_gmx-users" Cc:"ZHANG Cheng"<272699...@qq.com; > Subject:Simulate only one unit of the virus capsid while fixing > its surrounding units > > > > It is a challenge to simulate the entire virus as it is too big and I > do not have such computational resources. So I was thinking to only > simulate one coat protein and its surrounding neighbours, but keep the > neighbours relatively fixed. > > > Can I ask > > > 1) Is this a sensible idea to proceed? > > > 2) To fix the neighbours, should I use "constraints" or "restraints"? > > > 3) At which step should I start to introduce the fixation? > > > 4) If possible, is there a tutorial for this? I feel the information > here is still not straightforward to follow > http://www.gromacs.org/Documentation/How-tos/Position_Restraints > > > Thank you! > > > Yours sincerely > Cheng > -- > 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.
