Re: [gmx-users] Improving scaling - Gromacs 4.0 RC2
Justin A. Lemkul wrote: Hi, I've been playing around with the latest release candidate of version 4.0, and I was hoping someone out there more knowledgeable than me might tell me how to improve a bit on the performance I'm seeing. To clarify, the performance I'm seeing is a ton faster than 3.3.x, but I still seem to be getting bogged down with the PME/PP balance. I'm using mostly the default options with the new mdrun: mdrun_mpi -s test.tpr -np 64 -npme 32 try -dlb auto The system contains about 150,000 atoms - a membrane protein surrounded by several hundred lipids and solvent (water). The protein parameters are GROMOS, lipids are Berger, and water is SPC. My .mdp file (adapted from a generic 3.3.x file that I always used to use for such simulations) is attached at the end of this mail. It seems that my system runs fastest on 64 CPU's. Almost all tests with 128 or 256 seem to run slower. The nodes are dual-core 2.3 GHz Xserve G5, connected by Infiniband. Here's a summary of some of the tests I've run: -np-npme-ddorderns/day% performance loss from imbalance 6416interleave5.76019.6 6432interleave9.60040.9 6432pp_pme5.2523.9 6432cartesian5.3834.7 All other mdrun command line options are defaults. I get ~10.3 ns/day with -np 256 -npme 64, but since -np 64 -npme 32 seems to give almost that same performance there seems to be no compelling reason to tie up that many nodes. Any hints on how to speed things up any more? Is it possible? Not that I'm complaining...the same system under GMX 3.3.3 gives just under 1 ns/day :) I'm really curious about the 40.9% performance loss I'm seeing with -np 64 -npme 32, even though it gives the best overall performance in terms of ns/day. Thanks in advance for your attention, and any comments. -Justin ===test.mdp= title= NPT simulation for a membrane protein ; Run parameters integrator= md dt= 0.002 nsteps= 1; 20 ps nstcomm= 1 ; Output parameters nstxout= 500 nstvout= 500 nstfout= 500 nstlog= 500 nstenergy= 500 ; Bond parameters constraint_algorithm = lincs constraints= all-bonds continuation = no; starting up ; Twin-range cutoff scheme, parameters for Gromos96 nstlist= 5 ns_type= grid rlist= 0.8 rcoulomb= 0.8 rvdw= 1.4 ; PME electrostatics parameters coulombtype= PME fourierspacing = 0.24 pme_order= 4 ewald_rtol= 1e-5 optimize_fft= yes ; V-rescale temperature coupling is on in three groups Tcoupl = V-rescale tc_grps= Protein POPC SOL_NA+_CL- tau_t= 0.1 0.1 0.1 ref_t= 310 310 310 ; Pressure coupling is on Pcoupl= Berendsen pcoupltype= semiisotropic tau_p= 2.0 compressibility= 4.5e-5 4.5e-5 ref_p= 1.0 1.0 ; Generate velocities is on gen_vel= yes gen_temp= 310 gen_seed= 173529 ; Periodic boundary conditions are on in all directions pbc= xyz ; Long-range dispersion correction DispCorr= EnerPres end test.mdp== -- David. David van der Spoel, PhD, Professor of Biology Dept. of Cell and Molecular Biology, Uppsala University. Husargatan 3, Box 596, 75124 Uppsala, Sweden phone: 46 18 471 4205 fax: 46 18 511 755 [EMAIL PROTECTED] [EMAIL PROTECTED] http://folding.bmc.uu.se ___ gmx-users mailing listgmx-users@gromacs.org http://www.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
Re: [gmx-users] g_hbond, residue selection and boundary conditions
Hi Nicolas, 1. Does g_hbond takes into account the periodic boundary conditions or should first I center the system on my molecule, then run g_hbond? Yes (pretty much all tools do). 2. When I run g_hbond, it prints the message No option -sel the usual welcome message and run normally after that. I don't use any -sel option in the command line. Is it because I'm using g_hbond from gmx 3.3.3 with a trajectory from gmx 3.3.1? I use the following command line: It's an innocent bug in g_hbond, nothing to do with your command-line or system 3. When g_hbond ask for 2 groups, is it the same think to specify group1 then group2 and group2 then group1? (assuming group1 and group2 do not overlap) Yes. The number of hydrogen bonds between A and B is equal to the number between B and A. Cheers, Tsjerk -- Tsjerk A. Wassenaar, Ph.D. Junior UD (post-doc) Biomolecular NMR, Bijvoet Center Utrecht University Padualaan 8 3584 CH Utrecht The Netherlands P: +31-30-2539931 F: +31-30-2537623 ___ gmx-users mailing listgmx-users@gromacs.org http://www.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
Re: [gmx-users] rot_correlation
On Wed, 1 Oct 2008 12:22:03 -0400 rams rams [EMAIL PROTECTED] wrote: Hi Xavier, I started working on the g_rms program (to obtain the rotation matrices) suggested by you as well as it was mentioned in a couple of papers for the calcualtion of rotational correlation times of proteins. Meanwhile, I have an idea like the following : The problem in using the MD generated trajectory to obtain the rotational correlation times is, the vectors / the system will have all the three degrees of freedom here. (Though by using the option -fit rot+trans option in trjconv, we can remove the translational and rotational degrees of freedom but still the internal vibrations are there). Where as for obtaining the rotational correaltion times, we like to know how a particular vector or a system is reorienting during the course of the simulation alone. For this, if I superimpose my reference structure (.tpr) on the structures generated at different time scales (it is just opposite to the regular rms fits) and if I generate a trajectory with the rotated structures of .tpr files at various time intervals, this trajectory gives me the orientation of a vector / system at different time scales. Also the lenghts of the vectors are also the same (hence no need to worry about the internal motions). Can I use this trajectory to obtain the rotaitonal correlation times ?? If I understand correctly you like to generate a duplicate of your trajectory replacing each frame by a reference structure to remove the internal motions. You would then use this duplicate trajectory to get the time auto-correlation function of the NH vectors and extract the overall rotational correlation time of the protein. If this is correct it would be equivalent to use the rotation matrix from g_rms (or any other tool). Note that if you do not have specific experimental data to compare to it is not so important which method you use, but you have to explain how you get the number. Note also that g_dipole give the molecular dipole relaxation time which is (I think) what you are looking for. I hope this helps. XAvier. Though its lot of work to do the things manually, but i can give a try if it is alright to try in the above mentioned way. Thanks in advance. ram. - XAvier Periole - PhD - Molecular Dynamics Group - NMR and Computation University of Groningen The Netherlands - ___ gmx-users mailing listgmx-users@gromacs.org http://www.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
[gmx-users] error NH2 in C terminal
Dear Justin Thank you for your advice. there is NH2 in c terminal sequence peptide but when I do pdb2gmx I get error.I delete NH2 by hand and do pdb2gmx -f name.pdb -ter and choose none in c terminal .when I see top file I see NH2 add in C terminal.Is it correct? and another way I use missing in command pdb2gmx .I get top file similar to first top(according delet NH2).Please advise me,what is differnce top files with two way and which one is better I choose? thanks again -- sh-karbalaee ___ gmx-users mailing listgmx-users@gromacs.org http://www.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
Re: [gmx-users] Improving scaling - Gromacs 4.0 RC2
Hi Justin, I have written a small gmx tool that tries various PME/PP balances systematically for a given number of nodes and afterwards gives a suggestion what the fastest combindation is. Although I plan to extend it with more functionality, it's already working and I can send it to you if you like to try it. The performance loss due to load imbalance has two reasons: 1. imbalance in the force calculation, which can be levelled out by using -dlb yes or -dlb auto as David suggested 2. imbalance in short-range / long range force calculation which can be levelled by choosing the optimal PME/PP ratio. This the script should do for you. You might also want to check the md.log file for more detailed information about where your imbalance is coming from. My guess is that with 32 PME nodes and interleaved communication the PP (short range) nodes have to wait on the PME (long range) nodes, while with 16 PME nodes it is the other way around. That you see less imbalance with pp_pme or cartesian communication probably only means that the PME communication is slower in this case - the 'smaller' performance loss from imbalance is a bit misleading here. Carsten Am 01.10.2008 um 23:18 schrieb Justin A. Lemkul: Hi, I've been playing around with the latest release candidate of version 4.0, and I was hoping someone out there more knowledgeable than me might tell me how to improve a bit on the performance I'm seeing. To clarify, the performance I'm seeing is a ton faster than 3.3.x, but I still seem to be getting bogged down with the PME/ PP balance. I'm using mostly the default options with the new mdrun: mdrun_mpi -s test.tpr -np 64 -npme 32 The system contains about 150,000 atoms - a membrane protein surrounded by several hundred lipids and solvent (water). The protein parameters are GROMOS, lipids are Berger, and water is SPC. My .mdp file (adapted from a generic 3.3.x file that I always used to use for such simulations) is attached at the end of this mail. It seems that my system runs fastest on 64 CPU's. Almost all tests with 128 or 256 seem to run slower. The nodes are dual- core 2.3 GHz Xserve G5, connected by Infiniband. Here's a summary of some of the tests I've run: -np -npme -ddorderns/day % performance loss from imbalance 64 16 interleave 5.760 19.6 64 32 interleave 9.600 40.9 64 32 pp_pme 5.252 3.9 64 32 cartesian 5.383 4.7 All other mdrun command line options are defaults. I get ~10.3 ns/day with -np 256 -npme 64, but since -np 64 -npme 32 seems to give almost that same performance there seems to be no compelling reason to tie up that many nodes. Any hints on how to speed things up any more? Is it possible? Not that I'm complaining...the same system under GMX 3.3.3 gives just under 1 ns/day :) I'm really curious about the 40.9% performance loss I'm seeing with -np 64 -npme 32, even though it gives the best overall performance in terms of ns/day. Thanks in advance for your attention, and any comments. -Justin ===test.mdp= title = NPT simulation for a membrane protein ; Run parameters integrator = md dt = 0.002 nsteps = 1 ; 20 ps nstcomm = 1 ; Output parameters nstxout = 500 nstvout = 500 nstfout = 500 nstlog = 500 nstenergy = 500 ; Bond parameters constraint_algorithm= lincs constraints = all-bonds continuation= no; starting up ; Twin-range cutoff scheme, parameters for Gromos96 nstlist = 5 ns_type = grid rlist = 0.8 rcoulomb= 0.8 rvdw= 1.4 ; PME electrostatics parameters coulombtype = PME fourierspacing = 0.24 pme_order = 4 ewald_rtol = 1e-5 optimize_fft= yes ; V-rescale temperature coupling is on in three groups Tcoupl = V-rescale tc_grps = Protein POPC SOL_NA+_CL- tau_t = 0.1 0.1 0.1 ref_t = 310 310 310 ; Pressure coupling is on Pcoupl = Berendsen pcoupltype = semiisotropic tau_p = 2.0 compressibility = 4.5e-5 4.5e-5 ref_p = 1.0 1.0 ; Generate velocities is on gen_vel = yes gen_temp= 310 gen_seed= 173529 ; Periodic boundary conditions are on in all directions pbc = xyz ; Long-range dispersion correction DispCorr= EnerPres end test.mdp== -- Justin A. Lemkul Graduate Research Assistant Department of Biochemistry Virginia Tech Blacksburg, VA jalemkul[at]vt.edu | (540) 231-9080 http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the
RE: [gmx-users] Improving scaling - Gromacs 4.0 RC2
Hi, Looking at your 64 core results, it seems that your PP:PME load ratio is about 1:1. In most cases 3:1 is much better performance wise. grompp probably also printed a note about this and also how to fix it. I have also described this shortly in the parallelization section of the pdf manual. You should probably increase your cut-offs and pme grid spacing by the same factor (something around 1.2). Hopefully mdrun should choose the proper number of pme nodes for you when you do not use -npme. Berk Date: Wed, 1 Oct 2008 17:18:24 -0400 From: [EMAIL PROTECTED] To: gmx-users@gromacs.org Subject: [gmx-users] Improving scaling - Gromacs 4.0 RC2 Hi, I've been playing around with the latest release candidate of version 4.0, and I was hoping someone out there more knowledgeable than me might tell me how to improve a bit on the performance I'm seeing. To clarify, the performance I'm seeing is a ton faster than 3.3.x, but I still seem to be getting bogged down with the PME/PP balance. I'm using mostly the default options with the new mdrun: mdrun_mpi -s test.tpr -np 64 -npme 32 The system contains about 150,000 atoms - a membrane protein surrounded by several hundred lipids and solvent (water). The protein parameters are GROMOS, lipids are Berger, and water is SPC. My .mdp file (adapted from a generic 3.3.x file that I always used to use for such simulations) is attached at the end of this mail. It seems that my system runs fastest on 64 CPU's. Almost all tests with 128 or 256 seem to run slower. The nodes are dual-core 2.3 GHz Xserve G5, connected by Infiniband. Here's a summary of some of the tests I've run: -np -npme -ddorderns/day % performance loss from imbalance 6416 interleave 5.760 19.6 6432 interleave 9.600 40.9 6432 pp_pme 5.252 3.9 6432 cartesian 5.383 4.7 All other mdrun command line options are defaults. I get ~10.3 ns/day with -np 256 -npme 64, but since -np 64 -npme 32 seems to give almost that same performance there seems to be no compelling reason to tie up that many nodes. Any hints on how to speed things up any more? Is it possible? Not that I'm complaining...the same system under GMX 3.3.3 gives just under 1 ns/day :) I'm really curious about the 40.9% performance loss I'm seeing with -np 64 -npme 32, even though it gives the best overall performance in terms of ns/day. Thanks in advance for your attention, and any comments. -Justin ===test.mdp= title = NPT simulation for a membrane protein ; Run parameters integrator= md dt= 0.002 nsteps= 1 ; 20 ps nstcomm = 1 ; Output parameters nstxout = 500 nstvout = 500 nstfout = 500 nstlog= 500 nstenergy = 500 ; Bond parameters constraint_algorithm = lincs constraints = all-bonds continuation = no; starting up ; Twin-range cutoff scheme, parameters for Gromos96 nstlist = 5 ns_type = grid rlist = 0.8 rcoulomb = 0.8 rvdw = 1.4 ; PME electrostatics parameters coulombtype = PME fourierspacing = 0.24 pme_order = 4 ewald_rtol= 1e-5 optimize_fft = yes ; V-rescale temperature coupling is on in three groups Tcoupl= V-rescale tc_grps = Protein POPC SOL_NA+_CL- tau_t = 0.1 0.1 0.1 ref_t = 310 310 310 ; Pressure coupling is on Pcoupl= Berendsen pcoupltype= semiisotropic tau_p = 2.0 compressibility = 4.5e-5 4.5e-5 ref_p = 1.0 1.0 ; Generate velocities is on gen_vel = yes gen_temp = 310 gen_seed = 173529 ; Periodic boundary conditions are on in all directions pbc = xyz ; Long-range dispersion correction DispCorr = EnerPres end test.mdp== -- Justin A. Lemkul Graduate Research Assistant Department of Biochemistry Virginia Tech Blacksburg, VA jalemkul[at]vt.edu | (540) 231-9080 http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin ___ gmx-users mailing listgmx-users@gromacs.org http://www.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 _ Express yourself instantly with MSN Messenger! Download today it's FREE!
[gmx-users] Error in T-coupling of 2 atoms in T-couplingGroups
Hi, I am a beginner of Gromacs 3.3.3. I have done position restrined grompp with -f pr.mdp, -r .gro (water box having charge -2e), -c .gro and -p .top (produced by genbox). The output is a .tpr file. Then to neutralize the solution with Na+ ion I used genion by -s .tpr -o .gro - pname NA+ -np 2 -g .log. Then in the .top file in [molecule] portion I have changed #mol of SOL by reducing it by 2 and added NA+ of 2 mol to match the coordinate with .gro. And also deleted the .tpr file. Upto this step everything was OK. In next step again position restrained grompp was done using .gro file containing 2 Na+ ions. But which showed a Fatal error -- Fatal error: 2 atoms are not part of any of the T-coupling groups. So .tpr file is not generated. I am not understanding what to do now. Please help as early as possible. Thanking you, CHITRITA DUTTA ROY. ___ gmx-users mailing listgmx-users@gromacs.org http://www.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
Re: [gmx-users] HB lifetime
Please see my comments below. Hi, The HB definitions and associated lifetimes are a bit arbitrary, so there' s always going to be some ambiguity here. That being said, the reason the integral of the HB correlation function C(t) isn't an ideal definition is that C(t) is only roughly exponential. Same argument goes for getting the lifetime from a fit to C(t), or looking for the time where C(t)=1/e, or similar simple approximations. I disagree. HB lifetime is only slightly dependent on the exact values of the geometric parameters, around the usual values of R(O...O)= 3.5 Angstrom angle(O...O-H)= 30 degrees, please see JCP 129, 84505 (a link to the abstract is given below). C(t) of a HB obeys the analytical solution of the reversible geminate recombination (see a short review in JCP 129), and so its tail follows a power law: C(t) ~ Keq*(D*t)^-3/2, which is indicative of a 3 dimensions diffusion. What Luzar recommends is to think about an equilibrium between bound and unbound molecules, so that they interact with a forward and a backward rate constant k and k'. k gives the forward rate, ie. the HB breaking rate, and k' gives the HB reformation rate... they are not equal due to the diffusion of unbound molecules away from the solvation shell. There are a few advantages of going this route, not the least of which is that you tend to get similar lifetimes regardless of small changes in the HB definition, and whether you use geometric or energetic criteria, etc. The reversible geminate recombination deals with the A+B --- C, here A=B=H2O C=(H2O)2, the bound water dimer. From a single fit to C(t) one receives the bimolecular forward backward rate constants, which are well defined. k' you suggest is an apparent unimolecular rate constant, which appears to be more suited for short times. Extracting these rate constants is a bit tricky (I usually do it by hand), but I guess gromacs has a scheme to do it... I haven't actually looked at it (though I really should!). I'd recommend some caution though, a scheme that works well for HB's between water molecules in bulk may need to be adjusted to properly model HB's between water and polar atoms. I have to disagree again. The A+B=C problem has an analytical solution. Technically, ones only need to know how to calculate an error-function and to solve a cubic equation, please see eq. 9, 10 at JCP 129. The geminate problem is robust in the sense that it describes C(t) of ANY 2 particles, as long as their behavior is controlled by diffusion, it describes the water pair, but should describe also, for example, liquid argon. For the second case, ofcourse, different rate constants are expected. One should NOT see JCP 129 as a proof that previous works were absolutly wrong ! Instead, it shows that the postulate by Luzar Chandler, that C(t) of water is controlled by diffusion, is right, and that with the analytical solution of the geminate problem one can understand some aspects of the water dimer. For example - what causes the activation energies of the forward backward rate constants to be about similar rather then being different by the strength of 1 HB? Hope I was clear. Omer Markovitch. ** a link to JCP 129, 84505 (2008) http://dx.doi.org/10.1063/1.2968608 ** supporting information includes a short trajectory movie ___ gmx-users mailing listgmx-users@gromacs.org http://www.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
Re: [gmx-users] Hydrogen Bond Lifetime
Please see also my reply under HB lifetime to Christopher Daub. I think that JCP 129, 84505 (2008) should be read when dealing with C(t) in general HB lifetimes in particular. ** link: http://dx.doi.org/10.1063/1.2968608 Omer Markovitch. ___ gmx-users mailing listgmx-users@gromacs.org http://www.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
Re: [gmx-users] Error in T-coupling of 2 atoms in T-couplingGroups
Chitrita Duttaroy wrote: Hi, I am a beginner of Gromacs 3.3.3. http://3.3.3. I have done position restrined grompp with -f pr.mdp, -r .gro (water box having charge -2e), -c .gro and -p .top (produced by genbox). The output is a .tpr file. Then to neutralize the solution with Na+ ion I used genion by -s .tpr -o .gro - pname NA+ -np 2 -g .log. Then in the .top file in [molecule] portion I have changed #mol of SOL by reducing it by 2 and added NA+ of 2 mol to match the coordinate with .gro. And also deleted the .tpr file. Upto this step everything was OK. In next step again position restrained grompp was done using .gro file containing 2 Na+ ions. But which showed a Fatal error -- Fatal error: 2 atoms are not part of any of the T-coupling groups. So .tpr file is not generated. I am not understanding what to do now. Please help as early as possible. I'm guessing those two atoms are the 2 NA+ (most likely). Every element of your system should be included in a T-coupling group. Your .mdp file probably has not assigned the NA+ anywhere (merge it with the solvent in a special group: SOL_NA+). -Justin Thanking you, CHITRITA DUTTA ROY. ___ gmx-users mailing listgmx-users@gromacs.org http://www.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 -- Justin A. Lemkul Graduate Research Assistant Department of Biochemistry Virginia Tech Blacksburg, VA jalemkul[at]vt.edu | (540) 231-9080 http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin ___ gmx-users mailing listgmx-users@gromacs.org http://www.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
Re: [gmx-users] error NH2 in C terminal
shahrbanoo karbalaee wrote: Dear Justin Thank you for your advice. there is NH2 in c terminal sequence peptide but when I do pdb2gmx I get error.I delete NH2 by hand and do pdb2gmx -f name.pdb -ter and choose none in c terminal .when I see top file I see NH2 add in C terminal.Is it correct? I don't understand how deleting NH2 is causing it to show up in the topology. and another way I use missing in command pdb2gmx .I get top file similar to first top(according delet NH2).Please advise me,what is differnce top files with two way and which one is better I choose? Never use a topology that has missing atoms. What might be better is if you could paste the end of your .pdb file (showing the last residue with NH2, and the corresponding pdb2gmx command lines, then any modifications you have tried. Right now I do not understand what you are doing. -Justin thanks again -- Justin A. Lemkul Graduate Research Assistant Department of Biochemistry Virginia Tech Blacksburg, VA jalemkul[at]vt.edu | (540) 231-9080 http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin ___ gmx-users mailing listgmx-users@gromacs.org http://www.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
[gmx-users] Re:Error in T-coupling of 2 atoms in T-couplingGroups
Hi Chitrita, may be you haven't generated index file for the solvent and newly added ions use make_ndx file all the best Sudheer. I am a beginner of Gromacs 3.3.3. I have done position restrined grompp with -f pr.mdp, -r .gro (water box having charge -2e), -c .gro and -p .top (produced by genbox). The output is a .tpr file. Then to neutralize the solution with Na+ ion I used genion by -s .tpr -o .gro - pname NA+ -np 2 -g .log. Then in the .top file in [molecule] portion I have changed #mol of SOL by reducing it by 2 and added NA+ of 2 mol to match the coordinate with .gro. And also deleted the .tpr file. Upto this step everything was OK. In next step again position restrained grompp was done using .gro file containing 2 Na+ ions. But which showed a Fatal error -- Fatal error: 2 atoms are not part of any of the T-coupling groups. So .tpr file is not generated. I am not understanding what to do now. Please help as early as possible. Thanking you, ___ gmx-users mailing listgmx-users@gromacs.org http://www.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
[gmx-users] Uniform neutralizing plasma for Particle Mesh Ewald (PME) instead of counterions
Is there an implementation in gromacs for using a uniform neutralizing plasma with Particle Mesh Ewald (PME), to avoid use of counterions? Thank you -Himanshu ___ gmx-users mailing listgmx-users@gromacs.org http://www.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
Re: [gmx-users] Improving scaling - Gromacs 4.0 RC2
Carsten Kutzner wrote: Hi Justin, I have written a small gmx tool that tries various PME/PP balances systematically for a given number of nodes and afterwards gives a suggestion what the fastest combindation is. Although I plan to extend it with more functionality, it's already working and I can send it to you if you like to try it. I would like to try it; I think that it would help me get an empirical feel for things. The performance loss due to load imbalance has two reasons: 1. imbalance in the force calculation, which can be levelled out by using -dlb yes or -dlb auto as David suggested Indeed, using -dlb yes improves the % imbalance, but at the cost of output. If we take my 64-core case, applying -dlb yes reduces my speed to around 5-6 ns/day (depending on the # of PME nodes - 16, 24, 32). It just seems peculiar to me that I can get great speed in terms of ns/day using 64/32 PP/PME, but performance is hampered due to imbalance. 2. imbalance in short-range / long range force calculation which can be levelled by choosing the optimal PME/PP ratio. This the script should do for you. You might also want to check the md.log file for more detailed information about where your imbalance is coming from. My guess is that with 32 PME nodes and interleaved communication the PP (short range) nodes have to wait on the PME (long range) nodes, while with 16 PME nodes it is the other way around. That you see less imbalance with pp_pme or cartesian communication probably only means that the PME communication is slower in this case - the 'smaller' performance loss from imbalance is a bit misleading here. Ah, that makes a bit more sense. Thanks :) -Justin Carsten Am 01.10.2008 um 23:18 schrieb Justin A. Lemkul: Hi, I've been playing around with the latest release candidate of version 4.0, and I was hoping someone out there more knowledgeable than me might tell me how to improve a bit on the performance I'm seeing. To clarify, the performance I'm seeing is a ton faster than 3.3.x, but I still seem to be getting bogged down with the PME/PP balance. I'm using mostly the default options with the new mdrun: mdrun_mpi -s test.tpr -np 64 -npme 32 The system contains about 150,000 atoms - a membrane protein surrounded by several hundred lipids and solvent (water). The protein parameters are GROMOS, lipids are Berger, and water is SPC. My .mdp file (adapted from a generic 3.3.x file that I always used to use for such simulations) is attached at the end of this mail. It seems that my system runs fastest on 64 CPU's. Almost all tests with 128 or 256 seem to run slower. The nodes are dual-core 2.3 GHz Xserve G5, connected by Infiniband. Here's a summary of some of the tests I've run: -np-npme-ddorderns/day% performance loss from imbalance 6416interleave5.76019.6 6432interleave9.60040.9 6432pp_pme5.2523.9 6432cartesian5.3834.7 All other mdrun command line options are defaults. I get ~10.3 ns/day with -np 256 -npme 64, but since -np 64 -npme 32 seems to give almost that same performance there seems to be no compelling reason to tie up that many nodes. Any hints on how to speed things up any more? Is it possible? Not that I'm complaining...the same system under GMX 3.3.3 gives just under 1 ns/day :) I'm really curious about the 40.9% performance loss I'm seeing with -np 64 -npme 32, even though it gives the best overall performance in terms of ns/day. Thanks in advance for your attention, and any comments. -Justin ===test.mdp= title= NPT simulation for a membrane protein ; Run parameters integrator= md dt= 0.002 nsteps= 1; 20 ps nstcomm= 1 ; Output parameters nstxout= 500 nstvout= 500 nstfout= 500 nstlog= 500 nstenergy= 500 ; Bond parameters constraint_algorithm = lincs constraints= all-bonds continuation = no; starting up ; Twin-range cutoff scheme, parameters for Gromos96 nstlist= 5 ns_type= grid rlist= 0.8 rcoulomb= 0.8 rvdw= 1.4 ; PME electrostatics parameters coulombtype= PME fourierspacing = 0.24 pme_order= 4 ewald_rtol= 1e-5 optimize_fft= yes ; V-rescale temperature coupling is on in three groups Tcoupl = V-rescale tc_grps= Protein POPC SOL_NA+_CL- tau_t= 0.1 0.1 0.1 ref_t= 310 310 310 ; Pressure coupling is on Pcoupl= Berendsen pcoupltype= semiisotropic tau_p= 2.0 compressibility= 4.5e-5 4.5e-5 ref_p= 1.0 1.0 ; Generate velocities is on gen_vel= yes gen_temp= 310 gen_seed= 173529 ; Periodic boundary conditions are on in all directions pbc= xyz ; Long-range dispersion correction DispCorr= EnerPres end test.mdp== --
Re: [gmx-users] Improving scaling - Gromacs 4.0 RC2
Berk Hess wrote: Hi, Looking at your 64 core results, it seems that your PP:PME load ratio is about 1:1. In most cases 3:1 is much better performance wise. grompp probably also printed a note about this and also how to fix it. From the .mdp file I posted before, grompp gave the following: Calculating fourier grid dimensions for X Y Z Using a fourier grid of 56x60x50, spacing 0.238 0.230 0.240 Estimate for the relative computational load of the PME mesh part: 0.34 Should it have advised me about anything else? It seems that the PME load is reasonable, given what I understand about the matter. I suppose that does indicate to the PP/PME ratio I should be using. I have also described this shortly in the parallelization section of the pdf manual. You should probably increase your cut-offs and pme grid spacing by the same factor (something around 1.2). Which cut-offs, rlist/rcoulomb? I thought these were force field-dependent. Please correct me if I'm wrong. Hopefully mdrun should choose the proper number of pme nodes for you when you do not use -npme. I have never gotten mdrun to cooperate without specifically defining -npme; maybe it's just me or something that I'm doing. For example, output from two runs I tried (using 64 cores): 1. With fourierspacing = 0.12 (only difference from the posted .mdp file) --- Program mdrun_4.0_rc2_mpi, VERSION 4.0_rc2 Source code file: domdec_setup.c, line: 132 Fatal error: Could not find an appropriate number of separate PME nodes. i.e. = 0.563840*#nodes (34) and = #nodes/2 (32) and reasonable performance wise (grid_x=112, grid_y=117). Use the -npme option of mdrun or change the number of processors or the PME grid dimensions, see the manual for details. --- 2. With fourierspacing = 0.24 (the posted .mdp file) --- Program mdrun_4.0_rc2_mpi, VERSION 4.0_rc2 Source code file: domdec_setup.c, line: 132 Fatal error: Could not find an appropriate number of separate PME nodes. i.e. = 0.397050*#nodes (24) and = #nodes/2 (32) and reasonable performance wise (grid_x=56, grid_y=60). Use the -npme option of mdrun or change the number of processors or the PME grid dimensions, see the manual for details. --- Thanks. -Justin Berk Date: Wed, 1 Oct 2008 17:18:24 -0400 From: [EMAIL PROTECTED] To: gmx-users@gromacs.org Subject: [gmx-users] Improving scaling - Gromacs 4.0 RC2 Hi, I've been playing around with the latest release candidate of version 4.0, and I was hoping someone out there more knowledgeable than me might tell me how to improve a bit on the performance I'm seeing. To clarify, the performance I'm seeing is a ton faster than 3.3.x, but I still seem to be getting bogged down with the PME/PP balance. I'm using mostly the default options with the new mdrun: mdrun_mpi -s test.tpr -np 64 -npme 32 The system contains about 150,000 atoms - a membrane protein surrounded by several hundred lipids and solvent (water). The protein parameters are GROMOS, lipids are Berger, and water is SPC. My .mdp file (adapted from a generic 3.3.x file that I always used to use for such simulations) is attached at the end of this mail. It seems that my system runs fastest on 64 CPU's. Almost all tests with 128 or 256 seem to run slower. The nodes are dual-core 2.3 GHz Xserve G5, connected by Infiniband. Here's a summary of some of the tests I've run: -np -npme -ddorder ns/day % performance loss from imbalance 64 16 interleave 5.760 19.6 64 32 interleave 9.600 40.9 64 32 pp_pme 5.252 3.9 64 32 cartesian 5.383 4.7 All other mdrun command line options are defaults. I get ~10.3 ns/day with -np 256 -npme 64, but since -np 64 -npme 32 seems to give almost that same performance there seems to be no compelling reason to tie up that many nodes. Any hints on how to speed things up any more? Is it possible? Not that I'm complaining...the same system under GMX 3.3.3 gives just under 1 ns/day :) I'm really curious about the 40.9% performance loss I'm seeing with -np 64 -npme 32, even though it gives the best overall performance in terms of ns/day. Thanks in advance for your attention, and any comments. -Justin ===test.mdp= title = NPT simulation for a membrane protein ; Run parameters integrator = md dt = 0.002 nsteps = 1 ; 20 ps nstcomm = 1 ; Output parameters nstxout = 500 nstvout = 500 nstfout = 500 nstlog = 500 nstenergy = 500 ; Bond parameters constraint_algorithm = lincs constraints = all-bonds continuation = no ; starting up ; Twin-range cutoff scheme, parameters for Gromos96 nstlist = 5 ns_type = grid rlist = 0.8
[gmx-users] X2top
Dear gmxusers I want to make a itp file by x2top command. but unfortunately It does not work. I receive the following error: Fatal error: Library file ffoplsaa.n2t not found in current dir nor in default directories. (You can set the directories to search with the GMXLIB path variable) I check the root and also the files which this program could not find but I could not find the problem. What could the problem? thanks ___ gmx-users mailing listgmx-users@gromacs.org http://www.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
Re: [gmx-users] x2top
Morteza Khabiri wrote: Dear Dr,van der Spoel I am using the 3.3.2 version. It is totally installed. I used echo $GMXDATA but it did not show anything. Then you should run source /path/to/gromacsbin/GMXRC Thanks for your help Morteza ___ gmx-users mailing listgmx-users@gromacs.org http://www.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 -- David van der Spoel, Ph.D., Professor of Biology Molec. Biophys. group, Dept. of Cell Molec. Biol., Uppsala University. Box 596, 75124 Uppsala, Sweden. Phone: +46184714205. Fax: +4618511755. [EMAIL PROTECTED] [EMAIL PROTECTED] http://folding.bmc.uu.se ___ gmx-users mailing listgmx-users@gromacs.org http://www.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
Re: [gmx-users] HB lifetime
Christopher Daub wrote: Hi Omer, We are aware of your work with Dr. Agmon, and I believe Dr. Luzar has spoken with him about it. I don't understand it enough to say much, but I don't think we have substantive disagreements with it. Of course, the questioner was asking about the implementation of the Luzar model in Gromacs, so I tried to explain some of the background of her ideas. Perhaps they'll implement your HB model in Gromacs 5... I would encourage anyone to contribute implementation of this algorithm to the current g_hbond code. Please get in touch with me off-list if you are interested. Cheers, Chris. On Oct 2, 2008, at 4:45 AM, Omer Markovitch wrote: Please see my comments below. Hi, The HB definitions and associated lifetimes are a bit arbitrary, so there' s always going to be some ambiguity here. That being said, the reason the integral of the HB correlation function C(t) isn't an ideal definition is that C(t) is only roughly exponential. Same argument goes for getting the lifetime from a fit to C(t), or looking for the time where C(t)=1/e, or similar simple approximations. I disagree. HB lifetime is only slightly dependent on the exact values of the geometric parameters, around the usual values of R(O...O)= 3.5 Angstrom angle(O...O-H)= 30 degrees, please see JCP 129, 84505 (a link to the abstract is given below). C(t) of a HB obeys the analytical solution of the reversible geminate recombination (see a short review in JCP 129), and so its tail follows a power law: C(t) ~ Keq*(D*t)^-3/2, which is indicative of a 3 dimensions diffusion. What Luzar recommends is to think about an equilibrium between bound and unbound molecules, so that they interact with a forward and a backward rate constant k and k'. k gives the forward rate, ie. the HB breaking rate, and k' gives the HB reformation rate... they are not equal due to the diffusion of unbound molecules away from the solvation shell. There are a few advantages of going this route, not the least of which is that you tend to get similar lifetimes regardless of small changes in the HB definition, and whether you use geometric or energetic criteria, etc. The reversible geminate recombination deals with the A+B --- C, here A=B=H2O C=(H2O)2, the bound water dimer. From a single fit to C(t) one receives the bimolecular forward backward rate constants, which are well defined. k' you suggest is an apparent unimolecular rate constant, which appears to be more suited for short times. Extracting these rate constants is a bit tricky (I usually do it by hand), but I guess gromacs has a scheme to do it... I haven't actually looked at it (though I really should!). I'd recommend some caution though, a scheme that works well for HB's between water molecules in bulk may need to be adjusted to properly model HB's between water and polar atoms. I have to disagree again. The A+B=C problem has an analytical solution. Technically, ones only need to know how to calculate an error-function and to solve a cubic equation, please see eq. 9, 10 at JCP 129. The geminate problem is robust in the sense that it describes C(t) of ANY 2 particles, as long as their behavior is controlled by diffusion, it describes the water pair, but should describe also, for example, liquid argon. For the second case, ofcourse, different rate constants are expected. One should NOT see JCP 129 as a proof that previous works were absolutly wrong ! Instead, it shows that the postulate by Luzar Chandler, that C(t) of water is controlled by diffusion, is right, and that with the analytical solution of the geminate problem one can understand some aspects of the water dimer. For example - what causes the activation energies of the forward backward rate constants to be about similar rather then being different by the strength of 1 HB? Hope I was clear. Omer Markovitch. ** a link to JCP 129, 84505 (2008) http://dx.doi.org/10.1063/1.2968608 ** supporting information includes a short trajectory movie ___ gmx-users mailing listgmx-users@gromacs.org http://www.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 -- David. David van der Spoel, PhD, Professor of Biology Dept. of Cell and Molecular Biology, Uppsala University. Husargatan 3, Box 596, 75124 Uppsala, Sweden phone: 46 18 471 4205 fax: 46 18 511 755 [EMAIL PROTECTED] [EMAIL PROTECTED] http://folding.bmc.uu.se
[gmx-users] posres problem in 4.0_rc2. Bug in grompp?
Hi, I have a system in which I apply a posres on one atom. During em or md I get an error that I never had before: --- Program mdrun, VERSION 4.0_rc2 Source code file: mtop_util.c, line: 642 Software inconsistency error: Position restraint coordinates are missing --- Grompp runs withouth errors. When stargting mdrun, I get the error. The error is most likely in grompp: When running an older CVS-grompp (from last July), mdrun(4.0) works fine. When using the 4.0-grompp with the 4.0-mdrun, the error occurs. Or do I miss something? cheers, Jochen -- Dr. Jochen Hub Max Planck Institute for Biophysical Chemistry Computational biomolecular dynamics group Am Fassberg 11 D-37077 Goettingen, Germany Email: jhub[at]gwdg.de Tel.: +49 (0)551 201-2312 ___ gmx-users mailing listgmx-users@gromacs.org http://www.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
[gmx-users] RE: X2top
I check the root and also the files which this program could not find but I could not find the problem. What could the problem? What does it mean that you checked root? Is this file present in the gromacs topology folder? Are the access right OK? I think the problem is not in gromacs but in your particular system. -- Vitaly V. Chaban School of Chemistry National University of Kharkiv Svoboda sq.,4, Kharkiv 61077, Ukraine email: [EMAIL PROTECTED] skype: vvchaban tel.: +38-097-8259698 ___ gmx-users mailing listgmx-users@gromacs.org http://www.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
Re: [gmx-users] Error in T-coupling of 2 atoms in T-couplingGroups
Thank you Justin and Sudheer. I had already made a solution to my problem, but didn't know whether it was right. To confirm it I want to discuss it with you. After doing grompp to add Na+ to the water box I have done make_ndx with .gro file. Then in the index file I merged the index no. of 2 [NA+] into the [SOL] group and deleted the [NA+] group. After that I could do the position restrained grompp with the file that 2 Na+ ions. I want to know is it a correct method or not. Thanking you in advance, with regards CHITRITA. ___ gmx-users mailing listgmx-users@gromacs.org http://www.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
[gmx-users] Heat Flux and Lambda
I have been reading through the mailing list and have read some discussion that heat flux can be indicated by the value of lambda. Can I calculate the amount of heat flux from the lambda value? Thanks, Andy Shelley ___ gmx-users mailing listgmx-users@gromacs.org http://www.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
[gmx-users] off topic: graduate school opportunity
For those interested in graduate school doing molecular dynamics and free energy simulations, please read on. To the rest, sorry that this is somewhat off topic, but I thought it might be of interest to some because of my past involvement on the mailing list. Anyway, I just wanted to mention that I recently started my own research group in the Chemistry Department at the University of New Orleans. My group will be continuing various free energy studies, including working on methods for more accurate prediction of binding free energies, small molecule solvation and solubility, etc. If you are interested in a graduate program in chemistry, or you know someone who might be, and this work sounds interesting, I urge you to contact me at this e-mail address. I'm looking for some graduate students to join my group here. We are accepting applications NOW for the spring semester, and through March or so for Fall 2009. You can get some information on the department at http://www.chem.uno.edu and on admissions specifically at http://www.chem.uno.edu/ChemistryDepartmentfolder/Information.html, and if you need more info, contact me or the selections committee ([EMAIL PROTECTED]). While I do not currently have RA support for students, I anticipate having some in the near future, and in the meantime anyone joining my group would be supported by a TA position. Thanks for your time. David Mobley, Ph.D. Assistant Professor of Chemistry University of New Orleans New Orleans, LA 70148 Office 504-280-6445 Fax 504-280-6860 ___ gmx-users mailing listgmx-users@gromacs.org http://www.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
Re: [gmx-users] charge changes in free energy calculations
Hi, As long as you end up with the same charge in the initial and final states, you should be OK. It's only if your total transformation involves a change of net charge that you need to worry. So you should be fine if you turn off the charges on D, change the LJ interactions, and turn back on the charges on E. If I were you, I probably would not turn off the charges entirely -- after all, D and E share many atoms. I would only turn off charges on the atoms that you are going to modify. David On Fri, Sep 26, 2008 at 2:14 AM, friendli [EMAIL PROTECTED] wrote: Dear all, I have a mutation free energy calculation from D(asp) to E(glu). The charge is not changed for the overall mutation. However, following Dr. David Molbey's suggesion, electrostatic and VDW interaction should be modified separately, so in the first step we need to turn off the charge from D(-1) to D(0). I learn from the mailing list that it is problematic to do FE calculations with different charges for initial and final states. So is that safe to turn off the charge for D and turn on the same charge for E in this case? If not, it that OK to perform mutation FE calculation in one step? or there is no safe way to handle this kind mutations, i.e. mutating charge groups, currently without special correction? thank a lot Qiang ___ gmx-users mailing listgmx-users@gromacs.org http://www.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 ___ gmx-users mailing listgmx-users@gromacs.org http://www.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