Re: [gmx-users] Regarding-virtual-sites-tutorial
Hello Justin, Thanks for the mail. I am interested in knowing whether the particle type assigned has any role in the way it is treated by the program. For both virtual sites and the regular atoms, interaction parameters and mass are assigned. So in Gromacs, does the particle type (A or V) has any role beyond these interaction parameters. Regards, Sujith. -- 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] Alchemical free energy calculations - right choice of lambda points? What else might I do wrong?
p.s The discussion about convergence is relevant only if Replica Exchange (H-REMD/H-PT) is used and there is a high energy barrier in the dihedral potential. If the energy barrier is not higher than ~4.5*k_B*T you are welcome to try a calculation with and without zeroing the dihedral term and see that you get the same result. Thanks, Best, Asaf Quoting asaffa...@post.tau.ac.il: You're welcome. Please note that improper dihedral term which relates between the dummy atoms and the rest of the molecule needs to be removed (zeroed) as is explained in Section II. Thanks. Best regards, Asaf Quoting Julian Zachmann frankjulian.zachm...@uab.cat: Thanks a lot for your answers!!! I was really sure that the bonded terms to the dummy atoms should be zero so it would have taken me months to find this mistake!! I'm now running the simulations with the lambda points like this: ; 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 fep_lambdas = 0. 0.0500 0.1000 0.1500 0.2000 0.2500 0.3000 0.3500 0.4000 0.4667 0.5333 0.6000 0.6500 0.7000 0.7500 0.8000 0.8500 0.9000 0.9500 1. coul_lambdas= 0. 0.1000 0.2000 0.3000 0.5000 0.7000 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. restraint_lambdas = 0. 0.1000 0.2000 0.3000 0.5000 0.7000 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. vdw_lambdas = 0. 0. 0. 0. 0. 0. 0. 0.1000 0.2500 0.4000 0.5000 0.6000 0.6500 0.7000 0.7500 0.8000 0.8500 0.9000 0.9500 1. bonded_lambdas = 0. 0. 0. 0. 0. 0. 0. 0.1000 0.2500 0.4000 0.5000 0.6000 0.6500 0.7000 0.7500 0.8000 0.8500 0.9000 0.9500 1. Let's hope it works! Once again, thank you a lot. And of course I will quote the articles! 2015-07-30 10:12 GMT+02:00 asaffa...@post.tau.ac.il: Dear Hannes, It means that in many cases dihedral terms do not need to be transformed = not be multiplied by lambda= remain constant. Removing them also reduces the computational efficiency. The alternative is to remove them as is often done (zero them). We have shown not zero for dihedrals and all the other cases explained in the previous email. Best regards, Asaf Quoting hannes.loeff...@stfc.ac.uk: The argument regards dihedral that was made towards me was about convergence. When you say _can_ remain constant, what would be the alternative? I would still think not zero for bonds and angles. That is really the crucial point here. From: gromacs.org_gmx-users-boun...@maillist.sys.kth.se [ gromacs.org_gmx-users-boun...@maillist.sys.kth.se] on behalf of asaffa...@post.tau.ac.il [asaffa...@post.tau.ac.il] Sent: 30 July 2015 08:45 To: gromacs.org_gmx-users@maillist.sys.kth.se Subject: Re: [gmx-users] Alchemical free energy calculations - right choice of lambda points? What else might I do wrong? Hi, It has been proven both analytically (http://arxiv.org/abs/1310.2112, Section II) and in simulations (http://arxiv.org/abs/1310.2112, Section IV and http://arxiv.org/abs/1501.01514) that the dihedral terms can remain constant in the transformation in relative free energy calculations. In the second of these articles the dihedral term has been removed in a transformation both in vacuum and water and gave the same free energy difference, which means that removing them is an unnecessary operation (the contributions cancel out in the Thermodynamic Cycle). If there is an energy barrier in the dihedral potential significantly higher than k_B*T and there are several valleys it may or may not need to be multiplied by a value between 0 and 1 to help with convergence (see http://arxiv.org/abs/1310.2112, Section VI). In addition any bonded potential function which relates between the decoupled (dummy) atoms and the rest of the system - harmonic or non-harmonic (quadratic or non-quadratic) can remain constant. Also, bond angle potentials of sub-molecules with coupled degrees of freedom such as the methyl group can remain constant even if they relate between the decoupled atoms and the rest of the molecule (http://arxiv.org/abs/1310.2112, Section II). Please cite these articles if you are using them. Thanks, Best regards, Asaf Quoting hannes.loeff...@stfc.ac.uk: Hi, you wouid _not_ set bonded terms to zero. That would mean that you have essentially a non-bonded end state allowing the atom to be everywhere (reading Stefan Boresch' papers from 1999 on this may be helpful). The dihedrals are debatable (a collaborator of mine thinks that having them zero may help in replicate methods or so). Our strategy with FESetup (http://www.hecbiosim.ac.uk/fesetup) is to copy those terms over from the state without the dummies. The mass-lambdas are best kept at one end state as
[gmx-users] MD run complete but ligand not showing up
Hello I am confused about my MD results.According to Justin's tutorial of protein-ligand MD, i performed short 1ns md on my protein-ligand system. After completion, when i opened up the md.gro and md.xtc in VMD, it is only showing up protein.Ligand is absent from box. Also, no .trr file is generated. When i tried to view hydrogen bond interactions, it gave fatal error that no topol.trr is found. I have also prepared movie.pdb and viewed it in pymol, but in that also ligand is not showing. Can anyone please tell me if this is a common error? do i need to insert the ligand gile manually? I can attach the files if required. Regards Suniba Sent from my iPhone -- 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] MD run complete but ligand not showing up
On 8/2/15 6:06 AM, su wrote: Hello I am confused about my MD results.According to Justin's tutorial of protein-ligand MD, i performed short 1ns md on my protein-ligand system. After completion, when i opened up the md.gro and md.xtc in VMD, it is only showing up protein.Ligand is absent from box. Also, no .trr file is Molecules can't disappear. Either you never added the ligand or your attempts to visualize it are using the wrong residue name. Open the coordinate file in a text editor; if the ligand isn't there then it never was. generated. When i tried to view hydrogen bond interactions, it gave fatal error that no topol.trr is found. I have also prepared movie.pdb and viewed You won't get a .trr unless you ask for one by setting nst[xvf]out 0. it in pymol, but in that also ligand is not showing. Can anyone please tell me if this is a common error? do i need to insert the ligand gile manually? I If you followed my tutorial, you should know the answer to that already :) can attach the files if required. Regards Suniba The mailing list does not accept attachments, but in this case no additional files are needed. -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.
Re: [gmx-users] Regarding-virtual-sites-tutorial
On 8/2/15 6:18 AM, sujithkakkat . wrote: Hello Justin, Thanks for the mail. I am interested in knowing whether the particle type assigned has any role in the way it is treated by the program. For both virtual sites and the regular atoms, interaction parameters and mass are assigned. So in Gromacs, does the particle type (A or V) has any role beyond these interaction parameters. Yes, particle types matter in terms of thermostats, updating routines, etc. -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.
