Re: [gmx-users] Replica Exchange MD on more than 64 processors
On Mon, 28 Dec 2009, David van der Spoel wrote: bharat v. adkar wrote: On Mon, 28 Dec 2009, Mark Abraham wrote: bharat v. adkar wrote: On Sun, 27 Dec 2009, Mark Abraham wrote: bharat v. adkar wrote: On Sun, 27 Dec 2009, Mark Abraham wrote: bharat v. adkar wrote: Dear all, I am trying to perform replica exchange MD (REMD) on a 'protein in water' system. I am following instructions given on wiki (How-Tos - REMD). I have to perform the REMD simulation with 35 different temperatures. As per advise on wiki, I equilibrated the system at respective temperatures (total of 35 equilibration simulations). After this I generated chk_0.tpr, chk_1.tpr, ..., chk_34.tpr files from the equilibrated structures. Now when I submit final job for REMD with following command-line, itgives some error: command line: mpiexec -np 70 mdrun -multi 35 -replex 1000 -schk_.tpr-v error msg: --- Program mdrun_mpi, VERSION 4.0.7 Source code file: ../../../SRC/src/gmxlib/smalloc.c, line: 179 Fatal error: Not enough memory. Failed to realloc 790760 bytes for nlist-jjnr, nlist-jjnr=0x9a400030 (called from file ../../../SRC/src/mdlib/ns.c, line 503) --- Thanx for Using GROMACS - Have a Nice Day : Cannot allocate memory Error on node 19, will try to stop all the nodes Halting parallel program mdrun_mpi on CPU 19 out of 70 *** The individual node on the cluster has 8GB of physical memory and 16GBof swap memory. Moreover, when logged onto the individual nodes, it shows more than 1GB of free memory, so there should be no problem with cluster memory. Also, the equilibration jobs for the same system are run on the same cluster without any problem. What I have observed by submitting different test jobs with varying number of processors (and no. of replicas, wherever necessary), that any jobwith total number of processors = 64, runs faithfully without any problem.As soon as total number of processors are more than 64, it gives the above error. I have tested this with 65 processors/65 replicas also. This sounds like you might be running on fewer physical CPUsthan you have available. If so, running multiple MPI processes perphysical CPU can lead to memory shortage conditions. I don't understand what you mean. Do you mean, there might be morethan 8 processes running per node (each node has 8 processors)? But that also does not seem to be the case, as SGE (sun grid engine) output shows only eight processes per node. 65 processes can't have 8 processes per node. why can't it have? as i said, there are 8 processors per node. what i have not mentioned is that how many nodes it is using. The jobs got distributed over 9 nodes. 8 of which corresponds to 64 processors + 1 processor from 9th node. OK, that's a full description. Your symptoms are indicative of someone making an error somewhere. Since GROMACS works over more than 64 processors elsewhere, the presumption is that you are doing something wrong or the machine is not set up in the way you think it is or should be. To get the most effective help, you need to be sure you're providing full information - else we can't tell which error you're making or (potentially) eliminate you as a source of error. Sorry for not being clear in statements. As far I can tell you, job distribution seems okay to me. It is 1 job per processor. Does non-REMD GROMACS run on more than 64 processors? Does your cluster support using more than 8 nodes in a run? Can you run an MPI Hello world application that prints the processor and node ID across more than 64 processors? Yes, the cluster supports runs with more than 8 nodes. I generated a system with 10 nm water box and submitted on 80 processors. It was running fine. It printed all 80 NODEIDs. Also showed me when the job will get over. bharat Mark bharat Mark I don't know what you mean by swap memory. Sorry, I meant cache memory.. bharat Mark System: Protein + water + Na ions (total 46878 atoms) Gromacs version: tested with both v4.0.5 and v4.0.7 compiled with:
Re: [gmx-users] convert B-factor
Hi Antonio, You can do something like: grep -v ^#@ rmsf1.xvg rmsf1.dat grep -v ^#@ rmsf2.xvg rmsf2.dat paste rmsf1.dat rmsf2.dat | awk '{print $4-$2}' difference.dat That will give you a file with the rmsf difference. You can use editconf to read such data into the b-factor field, although you may need to modify the file a bit. Check editconf -h for that. Hope it helps, Tsjerk On Mon, Dec 28, 2009 at 7:52 AM, AntonioLeung royaltr...@live.cn wrote: I know how to calculate, and have calculated the RMSF of two trajectories (of the same molecule), and I want to compare the two RMSFs. I want convert their discrepancy into B-factors. Can you tell me more detailed? -- Original -- From: Mark Abrahammark.abra...@anu.edu.au; Date: Mon, Dec 28, 2009 11:04 AM To: Discussion list for GROMACS usersgmx-users@gromacs.org; Subject: Re: [gmx-users] convert B-factor AntonioLeung wrote: Dear all, I want to convert the difference of two rmsf data sets into B-factor of a coordinate (to illustrate their difference by coloring the structure by B-factor), can anyone tell me how to do it? g_rmsf -h Mark -- gmx-users mailing list gmx-users@gromacs.org 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 gmx-users-requ...@gromacs.org. Can't post? Read http://www.gromacs.org/mailing_lists/users.php -- gmx-users mailing list gmx-us...@gromacs.org 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 gmx-users-requ...@gromacs.org. Can't post? Read http://www.gromacs.org/mailing_lists/users.php -- Tsjerk A. Wassenaar, Ph.D. Computational Chemist Medicinal Chemist Neuropharmacologist -- gmx-users mailing listgmx-users@gromacs.org 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 gmx-users-requ...@gromacs.org. Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] protein simulation
edmund lee wrote: Dear all, I am trying to do a simulation of protein OMPA. At the step grompp, it shows a fatal error stated Fatal error: Atomtype 'HC' not found! I tried to configure the error but i failed. So, hope that anyone can help me in this. There are various underlying causes. Doing some (more) GROMACS tutorial material is probably a good idea. Mark -- gmx-users mailing listgmx-users@gromacs.org 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 gmx-users-requ...@gromacs.org. Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] Replica Exchange MD on more than 64 processors
bharat v. adkar wrote: On Mon, 28 Dec 2009, David van der Spoel wrote: bharat v. adkar wrote: On Mon, 28 Dec 2009, Mark Abraham wrote: bharat v. adkar wrote: On Sun, 27 Dec 2009, Mark Abraham wrote: bharat v. adkar wrote: On Sun, 27 Dec 2009, Mark Abraham wrote: bharat v. adkar wrote: Dear all, I am trying to perform replica exchange MD (REMD) on a 'protein in water' system. I am following instructions given on wiki (How-Tos - REMD). I have to perform the REMD simulation with 35 different temperatures. As per advise on wiki, I equilibrated the system at respective temperatures (total of 35 equilibration simulations). After this I generated chk_0.tpr, chk_1.tpr, ..., chk_34.tpr files from the equilibrated structures. Now when I submit final job for REMD with following command-line, itgives some error: command line: mpiexec -np 70 mdrun -multi 35 -replex1000 -schk_.tpr-v error msg: --- Program mdrun_mpi, VERSION 4.0.7 Source code file: ../../../SRC/src/gmxlib/smalloc.c, line:179 Fatal error: Not enough memory. Failed to realloc 790760 bytes for nlist-jjnr, nlist-jjnr=0x9a400030 (called from file ../../../SRC/src/mdlib/ns.c, line 503) --- Thanx for Using GROMACS - Have a Nice Day : Cannot allocate memory Error on node 19, will try to stop all the nodes Halting parallel program mdrun_mpi on CPU 19 out of 70 *** The individual node on the cluster has 8GB of physical memory and 16GBof swap memory. Moreover, when logged onto the individualnodes, it shows more than 1GB of free memory, so there should be no problem with cluster memory. Also, the equilibration jobs for the same system are run on the same cluster without any problem. What I have observed by submitting different test jobs with varying number of processors (and no. of replicas, wherever necessary), that any jobwith total number of processors = 64, runs faithfully without any problem.As soon as total number of processors are more than 64, it gives the above error. I have tested this with 65 processors/65 replicas also. This sounds like you might be running on fewer physicalCPUsthan you have available. If so, running multiple MPIprocesses perphysical CPU can lead to memory shortageconditions. I don't understand what you mean. Do you mean, there mightbe morethan 8 processes running per node (each node has 8 processors)? But that also does not seem to be the case, as SGE (sun grid engine) output shows only eight processes per node. 65 processes can't have 8 processes per node. why can't it have? as i said, there are 8 processors per node. what ihave not mentioned is that how many nodes it is using. The jobs got distributed over 9 nodes. 8 of which corresponds to 64 processors + 1 processorfrom 9th node. OK, that's a full description. Your symptoms are indicative of someone making an error somewhere. Since GROMACS works over more than 64 processors elsewhere, the presumption is that you are doing something wrong or the machine is not set up in the way you think it is or should be. To get the most effective help, you need to be sure you're providing full information - else we can't tell which error you're making or (potentially) eliminate you as a source of error. Sorry for not being clear in statements. As far I can tell you, job distribution seems okay to me. It is 1 jobper processor. Does non-REMD GROMACS run on more than 64 processors? Does your cluster support using more than 8 nodes in a run? Can you run an MPI Hello world application that prints the processor and node ID across more than 64 processors? Yes, the cluster supports runs with more than 8 nodes. I generated a system with 10 nm water box and submitted on 80 processors. It was running fine. It printed all 80 NODEIDs. Also showed me when the job will get over. bharat Mark bharat Mark I don't know what you mean by swap memory. Sorry, I meant cache memory.. bharat Mark System: Protein + water + Na ions (total 46878 atoms) Gromacs version: tested with both v4.0.5 and v4.0.7
[gmx-users] conversion between harmonic bonds/angles and GROMOS96 bonds/angles
I am interested in converting between GROMOS96 bond/angle potentials and standard harmonic potentials. The GROMACS manual (version 4) suggests formulae for doing so (Eq. 4.38 and Eq. 4.53). However, I am having trouble understanding where these come from and find that they seem to give rather poor approximations for the conversions. It seems that much more accurate (but somewhat more complex) conversion expressions can be derived but I am wondering whether I do not fully understand the GROMOS96 bond potentials vs. standard harmonic potentials. Any insights? Thanks, Michael. -- gmx-users mailing listgmx-users@gromacs.org 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 gmx-users-requ...@gromacs.org. Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] trjconv -pbc: how to keep all parts of the system clustered together in PDB?
I thought I was sure -pbc cluster will work, but it doesn't :( trjconv get's stuck on an infinite loop while calculating center of mass. In an index.ndx I created a new group which I called CLUSTER, as Mark suggested (I used make_ndx), then I ran trjconv: trjconv -f 1Y2Elig_em.trr -s 1Y2Elig_em.tpr -o tmp.pdb -b 120 -e 125 -pbc cluster -n index.ndx What I get is an infinite loop: COM:2.784 1.968 3.409 iter = 1 Isq = 1840514.500 COM:4.175 5.905 3.409 iter = 2 Isq = 230100.828 COM:2.784 1.968 3.409 iter = 3 Isq = 1840514.500 COM:4.175 5.905 3.409 iter = 4 Isq = 230100.828 COM:2.784 1.968 3.409 iter = 5 Isq = 1840514.500 COM:4.175 5.905 3.409 iter = 6 Isq = 230100.828 COM:2.784 1.968 3.409 iter = 7 Isq = 1840514.500 COM:4.175 5.905 3.409 iter = 8 Isq = 230100.828 COM:2.784 1.968 3.409 iter = 9 Isq = 1840514.500 COM:4.175 5.905 3.409 iter = 10 Isq = 230100.828 ... which goes on forever... What am I doing wrong? (Should I attach the files?) Thank you for your time! Vis Dear GROMACS users and gurus, I am sorry if it's a stupid question...I'm fairly new GROMACS, and something is been driving me crazy. I have a protein, two metal ions, and inhibitor in my system. Somehow in some of the frames I can't keep all those pieces clustered compactly for some postprocessing, using trjconv for conversion of trr/xtc into PDB format: -pbc mol option of trjconv: metal ions are far from the rest of the protein. -pbc nojump or -pbc whole: inhibitor far from the protein, metals are fine. -pbc atom or -pbc res: couple of residues are disattached from the protein; metals and inhibitor are fine. -pbc cluster: doesn't work (irrelevant?) -pbc cluster should work with a suitable index group of protein+metal+inhibitor - that's what it is for. Once that's done, you may want to re-run trjconv to apply other effects. Two-pass processing is often necessary. Also -center and -boxcenter don't seem to help... Can anybody suggest some tricks? I used octahedral box for my runs. don't seem to help also doesn't help. We can't guess what it was about your inputs and outputs that was contrary to your hopes :-) Mark -- gmx-users mailing listgmx-users@gromacs.org 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 gmx-users-requ...@gromacs.org. Can't post? Read http://www.gromacs.org/mailing_lists/users.php
[gmx-users] Problems with umbrella sampling
Hi, my system is constructed of 2 parallel plates in a box with solvent. each plate is made of 36 atoms which are positively charged on one of them (SRP) and negatively charged on the other plate (SRN). i want to perform a simulation with umbrella sampling between the 2 so i defined the pulling section parameters in the .mdp parameter file as following: ; COM PULLING ; Pull type: no, umbrella, constraint or constant_force pull = umbrella ; Pull geometry: distance, direction, cylinder or position pull_geometry= distance ; Select components for the pull vector. default: Y Y Y pull_dim = Y Y Y ; Cylinder radius for dynamic reaction force groups (nm) pull_r1 = 1 ; Switch from r1 to r0 in case of dynamic reaction force pull_r0 = 1.5 pull_constr_tol = 1e-06 pull_start = no pull_nstxout = 10 pull_nstfout = 1 ; Number of pull groups pull_ngroups = 2 ; Group name, weight (default all 1), vector, init, rate (nm/ps), kJ/(mol*nm^2) pull_group0 = SRP pull_weights0= 1 pull_pbcatom0= 0 pull_group1 = SRN pull_weights1= 1 pull_pbcatom1= 0 pull_vec1= 0.0 1.0 0.0 pull_init1 = 1.5 pull_rate1 = 0 pull_k1 = 1 pull_kB1 = 0 when proccesing the file with grompp i get the following error: *Fatal error: Number of weights (1) for pull group 0 'SRP' does not match the number of atoms (36)* is someone recognizing my mistake? does someone has an experience with umbrella sampling in GROMACS? thanks' amir * * -- gmx-users mailing listgmx-users@gromacs.org 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 gmx-users-requ...@gromacs.org. Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] Problems with umbrella sampling
Amir Marcovitz wrote: Hi, my system is constructed of 2 parallel plates in a box with solvent. each plate is made of 36 atoms which are positively charged on one of them (SRP) and negatively charged on the other plate (SRN). i want to perform a simulation with umbrella sampling between the 2 so i defined the pulling section parameters in the .mdp parameter file as following: ; COM PULLING ; Pull type: no, umbrella, constraint or constant_force pull = umbrella ; Pull geometry: distance, direction, cylinder or position pull_geometry= distance ; Select components for the pull vector. default: Y Y Y pull_dim = Y Y Y ; Cylinder radius for dynamic reaction force groups (nm) pull_r1 = 1 ; Switch from r1 to r0 in case of dynamic reaction force pull_r0 = 1.5 pull_constr_tol = 1e-06 pull_start = no pull_nstxout = 10 pull_nstfout = 1 ; Number of pull groups pull_ngroups = 2 ; Group name, weight (default all 1), vector, init, rate (nm/ps), kJ/(mol*nm^2) pull_group0 = SRP pull_weights0= 1 pull_pbcatom0= 0 pull_group1 = SRN pull_weights1= 1 pull_pbcatom1= 0 pull_vec1= 0.0 1.0 0.0 pull_init1 = 1.5 pull_rate1 = 0 pull_k1 = 1 pull_kB1 = 0 when proccesing the file with grompp i get the following error: *Fatal error: Number of weights (1) for pull group 0 'SRP' does not match the number of atoms (36)* is someone recognizing my mistake? Please refer to the manual (manual.gromacs.org is quite handy), you will find: Optional relative weights which are multiplied with the masses of the atoms to give the total weight for the COM. The number should be 0, meaning all 1, or the number of atoms in the pull group. I also think your value for pull_ngroup is wrong. It appears you are pulling SRN with respect to SRP, so you only have one pull group, not two. -Justin does someone has an experience with umbrella sampling in GROMACS? thanks' amir * * -- Justin A. Lemkul Ph.D. Candidate ICTAS Doctoral Scholar MILES-IGERT Trainee 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://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 gmx-users-requ...@gromacs.org. Can't post? Read http://www.gromacs.org/mailing_lists/users.php
[gmx-users] trjconv -pbc: how to keep all parts of the system clustered together in PDB?
Search: trjconv pbc cluster on the gromacs mailing list and take a look at the first hit. Basically, you need to find a frame that *does* work with -pbc cluster and then make a new .tpr based on the clustered .gro and then run trjconv -pbc mol. Just ensure that this frame is as close to the start of your run as possible. Chris. I thought I was sure -pbc cluster will work, but it doesn't :( trjconv get's stuck on an infinite loop while calculating center of mass. In an index.ndx I created a new group which I called CLUSTER, as Mark suggested (I used make_ndx), then I ran trjconv: trjconv -f 1Y2Elig_em.trr -s 1Y2Elig_em.tpr -o tmp.pdb -b 120 -e 125 -pbc cluster -n index.ndx What I get is an infinite loop: COM:2.784 1.968 3.409 iter = 1 Isq = 1840514.500 COM:4.175 5.905 3.409 iter = 2 Isq = 230100.828 COM:2.784 1.968 3.409 iter = 3 Isq = 1840514.500 COM:4.175 5.905 3.409 iter = 4 Isq = 230100.828 COM:2.784 1.968 3.409 iter = 5 Isq = 1840514.500 COM:4.175 5.905 3.409 iter = 6 Isq = 230100.828 COM:2.784 1.968 3.409 iter = 7 Isq = 1840514.500 COM:4.175 5.905 3.409 iter = 8 Isq = 230100.828 COM:2.784 1.968 3.409 iter = 9 Isq = 1840514.500 COM:4.175 5.905 3.409 iter = 10 Isq = 230100.828 ... which goes on forever... What am I doing wrong? (Should I attach the files?) Thank you for your time! Vis Dear GROMACS users and gurus, I am sorry if it's a stupid question...I'm fairly new GROMACS, and something is been driving me crazy. I have a protein, two metal ions, and inhibitor in my system. Somehow in some of the frames I can't keep all those pieces clustered compactly for some postprocessing, using trjconv for conversion of trr/xtc into PDB format: -pbc mol option of trjconv: metal ions are far from the rest of the protein. -pbc nojump or -pbc whole: inhibitor far from the protein, metals are fine. -pbc atom or -pbc res: couple of residues are disattached from the protein; metals and inhibitor are fine. -pbc cluster: doesn't work (irrelevant?) -pbc cluster should work with a suitable index group of protein+metal+inhibitor - that's what it is for. Once that's done, you may want to re-run trjconv to apply other effects. Two-pass processing is often necessary. Also -center and -boxcenter don't seem to help... Can anybody suggest some tricks? I used octahedral box for my runs. don't seem to help also doesn't help. We can't guess what it was about your inputs and outputs that was contrary to your hopes :-) Mark -- gmx-users mailing listgmx-users@gromacs.org 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 gmx-users-requ...@gromacs.org. Can't post? Read http://www.gromacs.org/mailing_lists/users.php
[gmx-users] the code speed between Gromacs 3.0 and Gromacs 4.0
Dear gmx-users, Are there any comparisons of the code speed between Gromacs 3.0 and Gromacs 4.0. In my calculation, I got a speed about 3.5ns/day of a system have about 50,000 atoms, using 8 CPUs with Gromacs 3.3.1. In contrast, the speed can reach up to 7.3ns/day when switch to Gromacs 4.0.7 while the other conditions are keep the same. Is this result normal? Best regards, -- gmx-users mailing listgmx-users@gromacs.org 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 gmx-users-requ...@gromacs.org. Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] the code speed between Gromacs 3.0 and Gromacs 4.0
Dechang Li wrote: Dear gmx-users, Are there any comparisons of the code speed between Gromacs 3.0 and Gromacs 4.0. In my calculation, I got a speed about 3.5ns/day of a system have about 50,000 atoms, using 8 CPUs with Gromacs 3.3.1. In contrast, the speed can reach up to 7.3ns/day when switch to Gromacs 4.0.7 while the other conditions are keep the same. Is this result normal? I would suggest reading the Gromacs 4 paper. There are lots of benchmarks in there. Sounds to me like everything is normal. -Justin Best regards, -- Justin A. Lemkul Ph.D. Candidate ICTAS Doctoral Scholar MILES-IGERT Trainee 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://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 gmx-users-requ...@gromacs.org. Can't post? Read http://www.gromacs.org/mailing_lists/users.php