RE: [gmx-users] Overflow problem with test-particle insertion
Hi, You should use the release-4-5-patches branch. After git clone do: git branch --track release-4-5-patches origin/release-4-5-patches git checkout release-4-5-patches Berk > Date: Thu, 9 Sep 2010 19:38:00 -0400 > From: kd...@princeton.edu > To: gmx-users@gromacs.org > Subject: [gmx-users] Overflow problem with test-particle insertion > > Hi, > > Thank you for taking the time to fix the issue. I would be very > interested in testing out the modified code, but unfortunately I have > had some difficulties compiling the Gromacs code obtained straight > from GIT. In particular, I encounter the following error: > > ~/gromacs/src/tools/gmx_membed.c:1095: error: expected declaration > specifiers or ‘...’ before ‘gmx_global_stat_t’ > > If you are curious to see the log and CMake cache files from the > build, I sent them in a previous email to the list (though the text of > the message somehow got scrubbed). I could also just patch the > relevant parts of version 4.5.1 and test that out. If this would be > feasible, then what specific lines should I modify? > > -Kevin > > > Date: Thu, 9 Sep 2010 17:19:36 +0200 > > From: Berk Hess > > Subject: RE: [gmx-users] Overflow problem with test-particle insertion > > To: Discussion list for GROMACS users > > Message-ID: > > Content-Type: text/plain; charset="iso-8859-1" > > > > > > Hi, > > > > I realized now that this is an SSE issue. > > Normally you would get NAN (or is it INF?). That is treated correctly in > > the GROMACS TPI code. > > But in SSE a float "wraps around" when it overflows, which could, in very > > few cases, lead to a reasonably > > looking energy value (I check for very high and very low values). > > I found that you can check for overflows in SSE and committed a fix for > > 4.5.2. > > I also filled the first 10 points (up to r=0.02 nm) of the potential/force > > tables, these used to be zero. > > These values are only relevant for energy minimization or TPI with extreme > > atomic overlap. > > > > Berk > > > > From: g...@hotmail.com > > To: gmx-users@gromacs.org > > Subject: RE: [gmx-users] Overflow problem with test-particle insertion > > Date: Thu, 9 Sep 2010 09:39:42 +0200 > > > > > > > > > > > > > > > > > > Hi, > > > > This is an interesting issue. > > The chance is quite small that this happens, but maybe not negligible. > > In single precision the maximum a float can store is 2^127. > > This gives a minimum distance of (2^127)^-1/12 = 6.5e-4 nm. > > The chance of inserting a particle within this radius is dens*3e-10, > > where dens is the number of particles per nm^3. > > A typical density of LJ particles is 30 per nm^3, which leads to a chance > > of 1e-8. > > Such insertion numbers can be reached, so we probably have to worry about > > this. > > > > However, in your example the distance seems to be around 4e-3, which would > > give r^-12 = 6e28. This still fits in a float and should not cause problems. > > So we should make sure we understand what's going on here. > > Could you file a bugzilla with the files to reproduce this and which > > insertion > > is the problematic one? > > > > I so two possible solutions: > > Force tabulated potentials with TPI, this can currently be achieved by > > setting > > the environment variable GMX_FORCE_TABLES > > Or require double precision. > > But I think both solutions would lead to about 40% lower performance. > > > > Berk > > > > > > Date: Wed, 8 Sep 2010 21:16:46 -0400 > > From: kd...@princeton.edu > > To: gmx-users@gromacs.org > > Subject: [gmx-users] Overflow problem with test-particle insertion > > > > Hello Gromacs users, > > > > I sent a message to the list in June describing what appeared to be a float > > overflow issue with the energy calculation for test-particle insertions: > > http://lists.gromacs.org/pipermail/gmx-users/2010-June/052213.html. > > > > > > I have recently tried the test-particle insertion mode in Gromacs-4.5.1, > > and it seems the problem is still there. Does anyone know how to work > > around or fix this problem without using tabulated potentials? > > > > -Kevin > > > -- > gmx-users mailing listgmx-users@gromacs.org > http://lists.gromacs.org/mailman/listinfo/gmx-users > Please search the archive at > http://www.gromacs.org/Support/Mailing_Lists/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/Support/Mailing_Lists -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/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/Support/Mailing_Lists
RE: [gmx-users] Overflow problem with test-particle insertion
Hi,
I put in the negative energy check on purpose.
There can be highly negative energies, which should not happen.
So energies smaller than -50kT are ignored.
But maybe such energies could only occur due to the SSE overflows
and not during non-SSE computation. If that's the case this check could
be removed.
Berk
Date: Thu, 9 Sep 2010 22:52:28 +0200
From: j...@um.es
To: gmx-users@gromacs.org
Subject: Re: [gmx-users] Overflow problem with test-particle insertion
Hello.
Probably that will not be related with the original problem but that's
something that came to me when reviewing the tpi.c code:
In line 583 (and follow), it is stated:
if (epot != epot || epot*beta < bU_neg_limit)
{
if (debug)
{
fprintf(debug,"\n time %.3f, step %d:
non-finite energy %f, using exp(-bU)=0\n",t,step,epot);
}
embU = 0;
}
And maybe the condition (after ||) should be:
if (epot != epot || -epot*beta < bU_neg_limit)
Otherwise we could be making a huge number (>exp(50)) equal to 0
instead of making exp(-50)=0 as intended, isn't it?
Well, I know I could be a bit confused
Javier
El 09/09/10 17:19, Berk Hess escribió:
Hi,
I realized now that this is an SSE issue.
Normally you would get NAN (or is it INF?). That is treated correctly
in the GROMACS TPI code.
But in SSE a float "wraps around" when it overflows, which could, in
very few cases, lead to a reasonably
looking energy value (I check for very high and very low values).
I found that you can check for overflows in SSE and committed a fix for
4.5.2.
I also filled the first 10 points (up to r=0.02 nm) of the
potential/force tables, these used to be zero.
These values are only relevant for energy minimization or TPI with
extreme atomic overlap.
Berk
From: g...@hotmail.com
To: gmx-users@gromacs.org
Subject: RE: [gmx-users] Overflow problem with test-particle insertion
Date: Thu, 9 Sep 2010 09:39:42 +0200
Hi,
This is an interesting issue.
The chance is quite small that this happens, but maybe not negligible.
In single precision the maximum a float can store is 2^127.
This gives a minimum distance of (2^127)^-1/12 = 6.5e-4 nm.
The chance of inserting a particle within this radius is dens*3e-10,
where dens is the number of particles per nm^3.
A typical density of LJ particles is 30 per nm^3, which leads to a
chance of 1e-8.
Such insertion numbers can be reached, so we probably have to worry
about this.
However, in your example the distance seems to be around 4e-3, which
would
give r^-12 = 6e28. This still fits in a float and should not cause
problems.
So we should make sure we understand what's going on here.
Could you file a bugzilla with the files to reproduce this and which
insertion
is the problematic one?
I so two possible solutions:
Force tabulated potentials with TPI, this can currently be achieved by
setting
the environment variable GMX_FORCE_TABLES
Or require double precision.
But I think both solutions would lead to about 40% lower performance.
Berk
Date: Wed, 8 Sep 2010 21:16:46 -0400
From: kd...@princeton.edu
To: gmx-users@gromacs.org
Subject: [gmx-users] Overflow problem with test-particle insertion
Hello Gromacs users,
I sent a message to the list in June describing what appeared to be a
float overflow issue with the energy calculation for test-particle
insertions: http://lists.gromacs.org/pipermail/gmx-users/2010-June/052213.html.
I have recently tried the test-particle insertion mode in
Gromacs-4.5.1, and it seems the problem is still there. Does anyone
know how to work around or fix this problem without using tabulated
potentials?
-Kevin
-- 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/Support/Mailing_Lists/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/Support/Mailing_Lists
-- 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/Support/Mailing_Lists/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/Support/Mailing_Lists
--
Javier CEREZO BASTIDA
Estudiante de Doctorado
-
Dpto. Química-Física
Universidad de Murcia
30100 MURCIA (España)
Tlf.(+34)868887434
--
gmx-users mailing listgmx-users@gromacs.org
http://lists.gromacs.org/mailman/listinfo/gmx-users
Please search the archive at
http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
P
[gmx-users] strange jumping in the frames
I am running Metadynamics with gromacs, I saw a jump in the frame after the simulation was running for a while, I copy here part of the log file: Step Time Lambda 2055750041115.00.0 Energies (kJ/mol) AngleProper Dih. LJ-14 Coulomb-14LJ (SR) 3.54705e+032.44201e+031.44990e+031.94769e+042.63535e+04 Coulomb (SR) Coul. recip. PotentialKinetic En. Total Energy -2.09459e+05 -5.86065e+04 -2.14796e+055.09371e+04 -1.63859e+05 Conserved En.Temperature Pressure (bar) Cons. rmsd () -1.66948e+053.47266e+02 -2.03134e+022.74833e-05 DD step 2055 vol min/aver 0.846 load imb.: force 0.4% Step Time Lambda 205641120.00.0 Energies (kJ/mol) AngleProper Dih. LJ-14 Coulomb-14LJ (SR) 3.36185e+032.48355e+031.45350e+031.94714e+042.67211e+04 Coulomb (SR) Coul. recip. PotentialKinetic En. Total Energy -2.09297e+05 -5.86137e+04 -2.14420e+055.18801e+04 -1.62540e+05 Conserved En.Temperature Pressure (bar) Cons. rmsd () -1.67082e+053.53695e+027.73024e+012.66928e-05 DD step 20562499 vol min/aver 0.825 load imb.: force 0.6% Step Time Lambda 2056250041125.003910.0 Energies (kJ/mol) AngleProper Dih. LJ-14 Coulomb-14LJ (SR) 3.55493e+032.49077e+031.45219e+031.93942e+042.60391e+04 Coulomb (SR) Coul. recip. PotentialKinetic En. Total Energy -2.08596e+05 -5.86454e+04 -2.14310e+055.14416e+04 -1.62868e+05 Conserved En.Temperature Pressure (bar) Cons. rmsd () -1.67250e+053.50706e+02 -3.50552e+022.93757e-05 as you can see the time jumped from 41120.0 to 41125.00391 , while it has to be 41125.0 and then it continue the rest of simulation like this, Do you what is the problem ?!! Fahimeh -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/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/Support/Mailing_Lists
[gmx-users] question about the format of [atomtypes] in ffnonbond.itp
Hi all, I just realized the the format of defining the atom-name in [atomtypes] section in ffnonbond.ipt in different forcefields are different. such as 1. oplsaa.ff/ [ atomtypes ] ; name bond_type mass charge ptype sigma epsilon opls_001 C 6 12.01100 0.500 A 3.75000e-01 4.39320e-01 ; SIG 2. charmm27.ff (similar as amber99.ff and gromos53a6.ff) [ atomtypes ] ;name at.num mass charge ptype sigma epsilon C 6 12.01100 0.51 A 0.356359487256 0.46024 3. Martini FF [ atomtypes ] ; name mass charge ptype c6 c12 P5 72.0 0.000 A 0.0 0.0 I wonder how gromacs deals with these difference smartly? And if I want to write my force field files, what should I pay attention to this difference? best wishes, Baofu Qiao -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/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/Support/Mailing_Lists
Re: [gmx-users] question about the format of [atomtypes] in ffnonbond.itp
- Original Message - From: Baofu Qiao Date: Friday, September 10, 2010 18:38 Subject: [gmx-users] question about the format of [atomtypes] in ffnonbond.itp To: Discussion list for GROMACS users --- | > Hi all, > > I just realized the the format of defining the atom-name in [atomtypes] > section in ffnonbond.ipt in different forcefields are different. such as > 1. oplsaa.ff/ > [ atomtypes ] > ; name bond_typemasscharge ptype sigma epsilon > opls_001 C 6 12.01100 0.500 A3.75000e-01 > 4.39320e-01 ; SIG > > 2. charmm27.ff (similar as amber99.ff and gromos53a6.ff) > [ atomtypes ] > ;name at.num masscharge ptype sigma epsilon > C 6 12.011000.51A 0.356359487256 0.46024 > > 3. Martini FF > [ atomtypes ] > ; name mass charge ptype c6 c12 > P5 72.0 0.000 A 0.0 0.0 > > I wonder how gromacs deals with these difference smartly? And if I want to > write my force field files, what should I pay attention to this difference? grompp does a bunch of black-magic sscanf manipulation, inferring things from the number of entries on lines and whether they are numbers or whatever. You don't want to know :-) For design purposes, matching any of them is probably fine. charmm27 looks like a reasonable compromise between information and noise. I can't think of an application for grompp knowing an atomic number, but it can't hurt. 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/Support/Mailing_Lists/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/Support/Mailing_Lists
Re: [gmx-users] strange jumping in the frames
> as you can see the time jumped from > 41120.0 to 41125.00391 , > while it has to be 41125.0 and then it continue the rest of > simulation like this, > Do you what is the problem ?!! Probably no problem at all. Numbers that are readily represented in decimal notation, like a timestep of 0.002ps, need not have an exact representation in *binary* floating point. If they don't, then when you add up a large number of them, then you can accumulate rounding error. Probably, if you run grep "0\.0" *.log then you will see other times reported that have not rounded nicely. My best guess is that your simulation stopped at this point, and restarted from the step corresponding to 41125.004ps, and rounding error is doing the rest. 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/Support/Mailing_Lists/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/Support/Mailing_Lists
[gmx-users] Protein out of the box (not PBC problem)
Hi everyone, please I can't figure out how to do this: I have a trajectory with water. If I concatenate it, since I have a FAT32 file system, it's a file too large! So I'm trying to avoid this. For that, I extracted my protein (it's a dimer) from each step of my simulation (when doing that, I did -pbc nojump and then I fit) and then concatenated all the steps, to obtain the whole trajectory with only the protein. When I watch my trajectory with VMD, I suspect that the distance between the monomers is larger than the dimension of my box. I'm sure that it's not a PBC problem, because I already did trjconv -pbc nojump. So my question is: is there a way to check if my protein is coming out of the box, without having to watch the whole trajectory with water molecules, because I can't do that. Thanks, Carla -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/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/Support/Mailing_Lists
Re: [gmx-users] Protein out of the box (not PBC problem)
- Original Message - From: Carla Jamous Date: Friday, September 10, 2010 19:15 Subject: [gmx-users] Protein out of the box (not PBC problem) To: Discussion list for GROMACS users > Hi everyone, > > please I can't figure out how to do this: > > I have a trajectory with water. If I concatenate it, since I have a FAT32 > file system, it's a file too large! So I'm trying to avoid this. > > For that, I extracted my protein (it's a dimer) from each step of my > simulation (when doing that, I did -pbc nojump and then I fit) and then > concatenated all the steps, to obtain the whole trajectory with only the > protein. When I watch my trajectory with VMD, I suspect that the distance > between the monomers is larger than the dimension of my box. > I'm sure that it's not a PBC problem, because I already did trjconv -pbc > nojump. > So my question is: is there a way to check if my protein is coming out of the > box, without having to watch the whole trajectory with water molecules, > because I can't do that. If you still have a box defined, then VMD can show that. Open the Tk console and enter "pbc box". You can also get VMD to show some periodic images - probably in the Representations. Maybe you want -center, or -fit progressive at some stage. trjconv has a million options... something will probably render things how you'd like them :-) Check out trjconv -h, of course. 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/Support/Mailing_Lists/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/Support/Mailing_Lists
[gmx-users] g_wham: "Pull geometry direction_periodic not supported"
Hi gmx-users I finished a series of pulling runs using the "periodic_direction" option and when I run the g_wham analysis I received the following message. --- Program g_wham, VERSION 4.5 Source code file: gmx_wham.c, line: 1369 Fatal error: Pull geometry direction_periodic not supported For more information and tips for troubleshooting, please check the GROMACS website at http://www.gromacs.org/Documentation/Errors --- Is there any way to circumvent this or should I redo all my simulations with the option of "direction" instead "direction_periodic"? Bests eef ___ Eudes Eterno Fileti Centro de Ciências Naturais e Humanas Universidade Federal do ABC — CCNH Av. dos Estados, 5001 Santo André - SP - Brasil CEP 09210-971 http://fileti.ufabc.edu.br -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/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/Support/Mailing_Lists
Re: [gmx-users] g_wham: "Pull geometry direction_periodic not supported"
Eudes Fileti wrote: Hi gmx-users I finished a series of pulling runs using the "periodic_direction" option and when I run the g_wham analysis I received the following message. --- Program g_wham, VERSION 4.5 Source code file: gmx_wham.c, line: 1369 Fatal error: Pull geometry direction_periodic not supported For more information and tips for troubleshooting, please check the GROMACS website at http://www.gromacs.org/Documentation/Errors --- Is there any way to circumvent this or should I redo all my simulations with the option of "direction" instead "direction_periodic"? See the usage notes here: http://www.gromacs.org/Documentation/How-tos/Potential_of_Mean_Force Unfortunately, this information is not "officially" documented, i.e. in the g_wham description. -Justin Bests eef ___ Eudes Eterno Fileti Centro de Ciências Naturais e Humanas Universidade Federal do ABC — CCNH Av. dos Estados, 5001 Santo André - SP - Brasil CEP 09210-971 http://fileti.ufabc.edu.br -- 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/Support/Mailing_Lists/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/Support/Mailing_Lists
RE: [gmx-users] Overflow problem with test-particle insertion
Hi,
I should read my own code comments better (usually Gromacs has too little
comments, but I put one here).
The lower bound check was because of issues with tabulated interactions.
The problem was that the Gromacs generated tables (for e.g. LJ shift or PME)
were zero for r<0.02 nm.
Thus you could miss LJ and/or Coulomb repulsion at near overlap.
But since I fixed this issue in the previous commit, this lower bound check can
go out.
I removed it for 4.5.2.
Berk
From: g...@hotmail.com
To: gmx-users@gromacs.org
Subject: RE: [gmx-users] Overflow problem with test-particle insertion
Date: Fri, 10 Sep 2010 10:16:48 +0200
Hi,
I put in the negative energy check on purpose.
There can be highly negative energies, which should not happen.
So energies smaller than -50kT are ignored.
But maybe such energies could only occur due to the SSE overflows
and not during non-SSE computation. If that's the case this check could
be removed.
Berk
Date: Thu, 9 Sep 2010 22:52:28 +0200
From: j...@um.es
To: gmx-users@gromacs.org
Subject: Re: [gmx-users] Overflow problem with test-particle insertion
Hello.
Probably that will not be related with the original problem but that's
something that came to me when reviewing the tpi.c code:
In line 583 (and follow), it is stated:
if (epot != epot || epot*beta < bU_neg_limit)
{
if (debug)
{
fprintf(debug,"\n time %.3f, step %d:
non-finite energy %f, using exp(-bU)=0\n",t,step,epot);
}
embU = 0;
}
And maybe the condition (after ||) should be:
if (epot != epot || -epot*beta < bU_neg_limit)
Otherwise we could be making a huge number (>exp(50)) equal to 0
instead of making exp(-50)=0 as intended, isn't it?
Well, I know I could be a bit confused
Javier
El 09/09/10 17:19, Berk Hess escribió:
Hi,
I realized now that this is an SSE issue.
Normally you would get NAN (or is it INF?). That is treated correctly
in the GROMACS TPI code.
But in SSE a float "wraps around" when it overflows, which could, in
very few cases, lead to a reasonably
looking energy value (I check for very high and very low values).
I found that you can check for overflows in SSE and committed a fix for
4.5.2.
I also filled the first 10 points (up to r=0.02 nm) of the
potential/force tables, these used to be zero.
These values are only relevant for energy minimization or TPI with
extreme atomic overlap.
Berk
From: g...@hotmail.com
To: gmx-users@gromacs.org
Subject: RE: [gmx-users] Overflow problem with test-particle insertion
Date: Thu, 9 Sep 2010 09:39:42 +0200
Hi,
This is an interesting issue.
The chance is quite small that this happens, but maybe not negligible.
In single precision the maximum a float can store is 2^127.
This gives a minimum distance of (2^127)^-1/12 = 6.5e-4 nm.
The chance of inserting a particle within this radius is dens*3e-10,
where dens is the number of particles per nm^3.
A typical density of LJ particles is 30 per nm^3, which leads to a
chance of 1e-8.
Such insertion numbers can be reached, so we probably have to worry
about this.
However, in your example the distance seems to be around 4e-3, which
would
give r^-12 = 6e28. This still fits in a float and should not cause
problems.
So we should make sure we understand what's going on here.
Could you file a bugzilla with the files to reproduce this and which
insertion
is the problematic one?
I so two possible solutions:
Force tabulated potentials with TPI, this can currently be achieved by
setting
the environment variable GMX_FORCE_TABLES
Or require double precision.
But I think both solutions would lead to about 40% lower performance.
Berk
Date: Wed, 8 Sep 2010 21:16:46 -0400
From: kd...@princeton.edu
To: gmx-users@gromacs.org
Subject: [gmx-users] Overflow problem with test-particle insertion
Hello Gromacs users,
I sent a message to the list in June describing what appeared to be a
float overflow issue with the energy calculation for test-particle
insertions: http://lists.gromacs.org/pipermail/gmx-users/2010-June/052213.html.
I have recently tried the test-particle insertion mode in
Gromacs-4.5.1, and it seems the problem is still there. Does anyone
know how to work around or fix this problem without using tabulated
potentials?
-Kevin
-- 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/Support/Mailing_Lists/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/Support/Mailing_Lists
-- gmx-users mailing list gmx-users@gromacs.org
http://lists.gromacs.org/mailman/listinfo/gmx-users
Please s
[gmx-users] Checkpoint files not being written
Hi All, I don't know if this is a new (intended) feature or a bug, so I figured I'd ask :) I'm running version 4.5.1 to do some implicit solvent simulations. What I've noticed is that, no matter how long the simulation runs, I only ever get a checkpoint file at the end of the simulation, and thus the checkpoint interval (which I have left at the default value of 15) is ignored. All the simulations I've done are sufficiently long (at least one hour) such that they should have produced checkpoint files according to this interval. Could this issue be related to the introduction of the -cpnum flag, which is set to "no" by default? I've poked around a bit in the mdrun code, but I easily lose my place trying to trace what's going on. I've tried specifically adding -cpnum to my mdrun command line, but still I am not receiving any intermediate checkpoint files at the requested -cpt interval. -Justin -- 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/Support/Mailing_Lists/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/Support/Mailing_Lists
RE: [gmx-users] Checkpoint files not being written
Hi, Do you have no cut-off's and nstlist set to 0? That's the only case where I could see that checkpoints are missing. If that's the case, please file a bugzilla so we don't forget to fix this for 4.5.2. PS -cpnum has nothing to do with this. I guess it's the bCPT = assignment in md.c that only gets set at NS steps or the last step. Berk > Date: Fri, 10 Sep 2010 11:08:26 -0400 > From: jalem...@vt.edu > To: gmx-users@gromacs.org > Subject: [gmx-users] Checkpoint files not being written > > > Hi All, > > I don't know if this is a new (intended) feature or a bug, so I figured I'd > ask > :) I'm running version 4.5.1 to do some implicit solvent simulations. What > I've noticed is that, no matter how long the simulation runs, I only ever get > a > checkpoint file at the end of the simulation, and thus the checkpoint > interval > (which I have left at the default value of 15) is ignored. All the > simulations > I've done are sufficiently long (at least one hour) such that they should > have > produced checkpoint files according to this interval. > > Could this issue be related to the introduction of the -cpnum flag, which is > set > to "no" by default? I've poked around a bit in the mdrun code, but I easily > lose my place trying to trace what's going on. I've tried specifically > adding > -cpnum to my mdrun command line, but still I am not receiving any > intermediate > checkpoint files at the requested -cpt interval. > > -Justin > > -- > > > 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/Support/Mailing_Lists/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/Support/Mailing_Lists -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/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/Support/Mailing_Lists
Re: [gmx-users] Checkpoint files not being written
With explicit solvent simulations I am still producing checkpoints as normal, maybe it has something to do with the implicit solvent settings? Cheers Tom On 10/09/10 16:08, Justin A. Lemkul wrote: Hi All, I don't know if this is a new (intended) feature or a bug, so I figured I'd ask :) I'm running version 4.5.1 to do some implicit solvent simulations. What I've noticed is that, no matter how long the simulation runs, I only ever get a checkpoint file at the end of the simulation, and thus the checkpoint interval (which I have left at the default value of 15) is ignored. All the simulations I've done are sufficiently long (at least one hour) such that they should have produced checkpoint files according to this interval. Could this issue be related to the introduction of the -cpnum flag, which is set to "no" by default? I've poked around a bit in the mdrun code, but I easily lose my place trying to trace what's going on. I've tried specifically adding -cpnum to my mdrun command line, but still I am not receiving any intermediate checkpoint files at the requested -cpt interval. -Justin -- Dr Thomas Piggot University of Southampton, UK. -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/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/Support/Mailing_Lists
RE: [gmx-users] Checkpoint files not being written
Hi, I guess you have nstlist=0. Try the fix below. Berk diff --git a/src/kernel/md.c b/src/kernel/md.c index dd92d51..225a4bb 100644 --- a/src/kernel/md.c +++ b/src/kernel/md.c @@ -1859,7 +1859,7 @@ double do_md(FILE *fplog,t_commrec *cr,int nfile,const t_filenm fnm[], * or at the last step (but not when we do not want confout), * but never at the first step or with rerun. */ -bCPT = (((gs.set[eglsCHKPT] && bNS) || +bCPT = (((gs.set[eglsCHKPT] && (bNS || ir->nstlist == 0)) || (bLastStep && (Flags & MD_CONFOUT))) && step > ir->init_step && !bRerunMD); if (bCPT) From: g...@hotmail.com To: gmx-users@gromacs.org Subject: RE: [gmx-users] Checkpoint files not being written Date: Fri, 10 Sep 2010 17:58:33 +0200 Hi, Do you have no cut-off's and nstlist set to 0? That's the only case where I could see that checkpoints are missing. If that's the case, please file a bugzilla so we don't forget to fix this for 4.5.2. PS -cpnum has nothing to do with this. I guess it's the bCPT = assignment in md.c that only gets set at NS steps or the last step. Berk > Date: Fri, 10 Sep 2010 11:08:26 -0400 > From: jalem...@vt.edu > To: gmx-users@gromacs.org > Subject: [gmx-users] Checkpoint files not being written > > > Hi All, > > I don't know if this is a new (intended) feature or a bug, so I figured I'd > ask > :) I'm running version 4.5.1 to do some implicit solvent simulations. What > I've noticed is that, no matter how long the simulation runs, I only ever get > a > checkpoint file at the end of the simulation, and thus the checkpoint > interval > (which I have left at the default value of 15) is ignored. All the > simulations > I've done are sufficiently long (at least one hour) such that they should > have > produced checkpoint files according to this interval. > > Could this issue be related to the introduction of the -cpnum flag, which is > set > to "no" by default? I've poked around a bit in the mdrun code, but I easily > lose my place trying to trace what's going on. I've tried specifically > adding > -cpnum to my mdrun command line, but still I am not receiving any > intermediate > checkpoint files at the requested -cpt interval. > > -Justin > > -- > > > 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/Support/Mailing_Lists/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/Support/Mailing_Lists -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/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/Support/Mailing_Lists -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/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/Support/Mailing_Lists
Re: [gmx-users] Checkpoint files not being written
Berk Hess wrote: Hi, I guess you have nstlist=0. Try the fix below. Yes, that is the case. I was using settings to employ the accelerated all-vs-all kernel. The fix below works and I have produced a .cpt file at the appropriate time. Thanks for the quick fix. -Justin Berk diff --git a/src/kernel/md.c b/src/kernel/md.c index dd92d51..225a4bb 100644 --- a/src/kernel/md.c +++ b/src/kernel/md.c @@ -1859,7 +1859,7 @@ double do_md(FILE *fplog,t_commrec *cr,int nfile,const t_filenm fnm[], * or at the last step (but not when we do not want confout), * but never at the first step or with rerun. */ -bCPT = (((gs.set[eglsCHKPT] && bNS) || +bCPT = (((gs.set[eglsCHKPT] && (bNS || ir->nstlist == 0)) || (bLastStep && (Flags & MD_CONFOUT))) && step > ir->init_step && !bRerunMD); if (bCPT) From: g...@hotmail.com To: gmx-users@gromacs.org Subject: RE: [gmx-users] Checkpoint files not being written Date: Fri, 10 Sep 2010 17:58:33 +0200 Hi, Do you have no cut-off's and nstlist set to 0? That's the only case where I could see that checkpoints are missing. If that's the case, please file a bugzilla so we don't forget to fix this for 4.5.2. PS -cpnum has nothing to do with this. I guess it's the bCPT = assignment in md.c that only gets set at NS steps or the last step. Berk > Date: Fri, 10 Sep 2010 11:08:26 -0400 > From: jalem...@vt.edu > To: gmx-users@gromacs.org > Subject: [gmx-users] Checkpoint files not being written > > > Hi All, > > I don't know if this is a new (intended) feature or a bug, so I figured I'd ask > :) I'm running version 4.5.1 to do some implicit solvent simulations. What > I've noticed is that, no matter how long the simulation runs, I only ever get a > checkpoint file at the end of the simulation, and thus the checkpoint interval > (which I have left at the default value of 15) is ignored. All the simulations > I've done are sufficiently long (at least one hour) such that they should have > produced checkpoint files according to this interval. > > Could this issue be related to the introduction of the -cpnum flag, which is set > to "no" by default? I've poked around a bit in the mdrun code, but I easily > lose my place trying to trace what's going on. I've tried specifically adding > -cpnum to my mdrun command line, but still I am not receiving any intermediate > checkpoint files at the requested -cpt interval. > > -Justin > > -- > > > 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 list gmx-users@gromacs.org > http://lists.gromacs.org/mailman/listinfo/gmx-users > Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/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/Support/Mailing_Lists -- 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/Support/Mailing_Lists/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/Support/Mailing_Lists -- 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/Support/Mailing_Lists/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/Support/Mailing_Lists
[gmx-users] g_hbond solvent insertion problem
Hi all: I´m using gromacs version 4.0.2 and a box type truncated octahedron for the MD simulation. I need to analyze hydrogen bonds with solvent (water)insertion. I used the following command: g_hbond f *.xtc s *.tpr n *.ndx num num.xvg dist *.xvg ang *.xvg a 45 r 0.3 ins. I selected acceptor and donor grups and solvent grup. The problem is that the file num.xvg is the same with direct hydrogen bonds (without solvent) and the output insert.dat is empty. Why is this? Any help would be appreciated Carmen -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/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/Support/Mailing_Lists
Re: [gmx-users] g_hbond solvent insertion problem
ceste...@unsl.edu.ar wrote: Hi all: I´m using gromacs version 4.0.2 and a box type truncated octahedron for the MD simulation. I need to analyze hydrogen bonds with solvent (water)insertion. I used the following command: g_hbond –f *.xtc –s *.tpr –n *.ndx –num num.xvg –dist *.xvg –ang *.xvg –a 45 –r 0.3 –ins. I selected acceptor and donor grups and solvent grup. The problem is that the file num.xvg is the same with direct hydrogen bonds (without solvent) and the output insert.dat is empty. Why is this? I don't think the -ins option has been functional for some time, and (I presume) it is for this reason that it was removed prior to the 4.5 release. -Justin Any help would be appreciated Carmen -- 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/Support/Mailing_Lists/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/Support/Mailing_Lists
[gmx-users] PDB will not minimize?
Downloaded a pdb from pdb.org (1HNX) and stripped the PCY ligand and plugged it into pdb2gmx using AMBER03 and TIP3P (GROMACS 4.5). Ran through the rest of my process and I am having some issues. Here is my flow of commands: pdb2gmx -f "$base".pdb -o "$base".gro -p "$base".top I use AMBER03 and TIP3P editconf -bt cubic -f "$base".gro -o "$base".gro -c -d 1.5 genbox -cp "$base".gro -cs spc216.gro -o "$base"_b4ion.gro -p "$base".top grompp -f em.mdp -c "$base"_b4ion.gro -p "$base".top -o "$base"_b4ion.tpr genion -s "$base"_b4ion.tpr -o "$base"_b4em.gro -neutral -conc 0.001 -pname NA -nname CL -g "$base"_ion.log -p "$base".top Here I have some issues too, but it is intermittent. I have a charge of -563 but genion wants to add 580+ NA and one if any CL giving me a possitive charge in the system. My thought is that -neutral would give me a neutral system? grompp -f em.mdp -c "$base"_b4em.gro -p "$base".top -o "$base"_em.tpr mdrun -v -s "$base"_em.tpr -c "$base"_after_em.gro -g emlog.log I have set this up to be very thorough and it can run for 1000+ steps. My problem is it never gives me a good Fmax. I am always above 1e-4. grompp -f pr.mdp -c "$base"_after_em.gro -p "$base".top -o "$base"_pr.tpr mdrun -v -s "$base"_pr.tpr -o "$base"_pr.trr -e pr.edr -c "$base"_after_pr.gro -g prlog.log -cpi state_pr.cpt -cpo state_pr.cpt This is where I always fail. First I get pressure errors and LINCS errors. I don't know how to minimize my system any farther, and therefor continue my jobs. grompp -f md.mdp -c "$base"_after_pr.gro -p "$base".top -o "$base"_md.tpr mdrun -v -s "$base"_md.tpr -o "$base"_md.trr -c "$base"_after_pr.gro -g md.log -e md.edr -cpi state_md.cpt -cpo state_md.cpt I NEVER get to here I have stripped the MG and ZN. I have removed a significant portion of the protein/rna to limit its size. I have ran it at 1 fs, 2 fs, 4 fs. I have ran it with -heavyh. And I have tried every AMBER force field available in GROMACS 4.5. Any help would be much appreciated. TJ Mustard Email: musta...@onid.orst.edu -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/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/Support/Mailing_Lists
[gmx-users] please help, cannot compile gmx 4.5 openmm
I am doing, after compiling and installing the normal grmx 4.5: rm -fr CMakeCache.txt make clean export OPENMM_ROOT_DIR=/usr/local/openmm cmake -DGMX_OPENMM=ON .. make mdrun [ 1%] Building NVCC (Device) object src/kernel/gmx_gpu_utils/./gmx_gpu_utils_generated_memtestG80_core.cu.o [ 1%] Building NVCC (Device) object src/kernel/gmx_gpu_utils/./gmx_gpu_utils_generated_gmx_gpu_utils.cu.o Scanning dependencies of target gmx_gpu_utils Linking CXX shared library libgmx_gpu_utils.dylib Undefined symbols: "_gmx_strncasecmp", referenced from: is_supported_cuda_gpu(int, char*)in gmx_gpu_utils_generated_gmx_gpu_utils.cu.o is_supported_cuda_gpu(int, char*)in gmx_gpu_utils_generated_gmx_gpu_utils.cu.o is_supported_cuda_gpu(int, char*)in gmx_gpu_utils_generated_gmx_gpu_utils.cu.o is_supported_cuda_gpu(int, char*)in gmx_gpu_utils_generated_gmx_gpu_utils.cu.o is_supported_cuda_gpu(int, char*)in gmx_gpu_utils_generated_gmx_gpu_utils.cu.o is_supported_cuda_gpu(int, char*)in gmx_gpu_utils_generated_gmx_gpu_utils.cu.o is_supported_cuda_gpu(int, char*)in gmx_gpu_utils_generated_gmx_gpu_utils.cu.o is_supported_cuda_gpu(int, char*)in gmx_gpu_utils_generated_gmx_gpu_utils.cu.o "_debug", referenced from: is_supported_cuda_gpu(int, char*)in gmx_gpu_utils_generated_gmx_gpu_utils.cu.o do_timed_memtest(int, int)in gmx_gpu_utils_generated_gmx_gpu_utils.cu.o do_full_memtest(int) in gmx_gpu_utils_generated_gmx_gpu_utils.cu.o do_quick_memtest(int) in gmx_gpu_utils_generated_gmx_gpu_utils.cu.o "_trim", referenced from: is_supported_cuda_gpu(int, char*)in gmx_gpu_utils_generated_gmx_gpu_utils.cu.o is_supported_cuda_gpu(int, char*)in gmx_gpu_utils_generated_gmx_gpu_utils.cu.o is_supported_cuda_gpu(int, char*)in gmx_gpu_utils_generated_gmx_gpu_utils.cu.o is_supported_cuda_gpu(int, char*)in gmx_gpu_utils_generated_gmx_gpu_utils.cu.o is_supported_cuda_gpu(int, char*)in gmx_gpu_utils_generated_gmx_gpu_utils.cu.o is_supported_cuda_gpu(int, char*)in gmx_gpu_utils_generated_gmx_gpu_utils.cu.o is_supported_cuda_gpu(int, char*)in gmx_gpu_utils_generated_gmx_gpu_utils.cu.o is_supported_cuda_gpu(int, char*)in gmx_gpu_utils_generated_gmx_gpu_utils.cu.o ld: symbol(s) not found collect2: ld returned 1 exit status make[3]: *** [src/kernel/gmx_gpu_utils/libgmx_gpu_utils.dylib] Error 1 make[2]: *** [src/kernel/gmx_gpu_utils/CMakeFiles/gmx_gpu_utils.dir/all] Error 2 make[1]: *** [src/kernel/CMakeFiles/mdrun.dir/rule] Error 2 make: *** [mdrun] Error 2 -- Alan Wilter S. da Silva, D.Sc. - CCPN Research Associate Department of Biochemistry, University of Cambridge. 80 Tennis Court Road, Cambridge CB2 1GA, UK. >>http://www.bio.cam.ac.uk/~awd28<< -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/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/Support/Mailing_Lists
Re: [gmx-users] PDB will not minimize?
TJ Mustard wrote: Downloaded a pdb from pdb.org (1HNX) and stripped the PCY ligand and plugged it into pdb2gmx using AMBER03 and TIP3P (GROMACS 4.5). Ran through the rest of my process and I am having some issues. Here is my flow of commands: pdb2gmx -f "$base".pdb -o "$base".gro -p "$base".top I use AMBER03 and TIP3P You should have failed right here. 1HNX has missing atoms in a THR residue, so pdb2gmx would otherwise exit with a fatal error (I know because I tried). I assume this means you have re-built the appropriate sidechain atoms? If so, how did you do it? Did you obtain a sensible .pdb file as output from whatever process you used? editconf -bt cubic -f "$base".gro -o "$base".gro -c -d 1.5 genbox -cp "$base".gro -cs spc216.gro -o "$base"_b4ion.gro -p "$base".top grompp -f em.mdp -c "$base"_b4ion.gro -p "$base".top -o "$base"_b4ion.tpr genion -s "$base"_b4ion.tpr -o "$base"_b4em.gro -neutral -conc 0.001 -pname NA -nname CL -g "$base"_ion.log -p "$base".top Here I have some issues too, but it is intermittent. I have a charge of -563 but genion wants to add 580+ NA and one if any CL giving me a possitive charge in the system. My thought is that -neutral would give me a neutral system? Unfortunately, genion doesn't deal well with large numbers for some reason. If you know how many ions you need to neutralize, do it manually with -np or -nn. Also of interest: how did you get this net charge? When I processed 1HNX (after re-building missing atoms, I have a charge in excess of -1000. grompp -f em.mdp -c "$base"_b4em.gro -p "$base".top -o "$base"_em.tpr mdrun -v -s "$base"_em.tpr -c "$base"_after_em.gro -g emlog.log I have set this up to be very thorough and it can run for 1000+ steps. My problem is it never gives me a good Fmax. I am always above 1e-4. An Fmax of 0.0001 is extremely stringent and unlikely to be produced for most systems in the absence of exhaustive EM using double precision. Such minimization is usually unnecessary for most systems, unless you're doing normal modes. The fact that you are "above" 0.0001 is thus unsurprising. How far above are you? Hard numbers, please, and do remind everyone of what your EM settings are. In assessing the inability to minimize, you should look at the atom on which the maximum force resides. It is likely one of the first to be destabilized during any subsequent MD. What atoms are around it that might be pushing on it? Is there any nonsensical geometry that may be present? What also might be substantially more beneficial would be to minimize in vacuo before going through all the effort of adding solvent and ions and thus just slowing your whole process down by adding tons more atoms to an already enormous system. If that doesn't work, minimize each chain separately (painstaking, I know) in vacuo to see where the problem(s) may lie. -Justin grompp -f pr.mdp -c "$base"_after_em.gro -p "$base".top -o "$base"_pr.tpr mdrun -v -s "$base"_pr.tpr -o "$base"_pr.trr -e pr.edr -c "$base"_after_pr.gro -g prlog.log -cpi state_pr.cpt -cpo state_pr.cpt This is where I always fail. First I get pressure errors and LINCS errors. I don't know how to minimize my system any farther, and therefor continue my jobs. grompp -f md.mdp -c "$base"_after_pr.gro -p "$base".top -o "$base"_md.tpr mdrun -v -s "$base"_md.tpr -o "$base"_md.trr -c "$base"_after_pr.gro -g md.log -e md.edr -cpi state_md.cpt -cpo state_md.cpt I NEVER get to here I have stripped the MG and ZN. I have removed a significant portion of the protein/rna to limit its size. I have ran it at 1 fs, 2 fs, 4 fs. I have ran it with -heavyh. And I have tried every AMBER force field available in GROMACS 4.5. Any help would be much appreciated. TJ Mustard Email: musta...@onid.orst.edu -- 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/Support/Mailing_Lists/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/Support/Mailing_Lists
[gmx-users] OWT5 not found in Charmmff
I am running tip5p in charmm force field in gmx. In energy minimization step, it gives an error msg as below. I just simply added it in atomtypes.atp in charmm27.ff, but still it doesn't work. Does anyone know how to solve it? Many thanks, Yao --- > Program grompp, VERSION 4.5-beta3 > Source code file: toppush.c, line: 1166 > > Fatal error: > Atomtype OWT5 not found > For more information and tips for troubleshooting, please > check the GROMACS > website at http://www.gromacs.org/Documentation/Errors > -- > >> -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/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/Support/Mailing_Lists
Re: [gmx-users] PDB will not minimize?
My appologies I have errors all over this. On September 11, 2010 at 12:44 AM "Justin A. Lemkul" wrote: > > > TJ Mustard wrote: > > > > > > Downloaded a pdb from pdb.org (1HNX) and stripped the PCY ligand and > > plugged it into pdb2gmx using AMBER03 and TIP3P (GROMACS 4.5). Ran > > through the rest of my process and I am having some issues. > > > > > > > > Here is my flow of commands: > > > > > > > > pdb2gmx -f "$base".pdb -o "$base".gro -p "$base".top > > > > I use AMBER03 and TIP3P > > > > > > You should have failed right here. 1HNX has missing atoms in a THR residue, so > pdb2gmx would otherwise exit with a fatal error (I know because I tried). I > assume this means you have re-built the appropriate sidechain atoms? If so, how > did you do it? Did you obtain a sensible .pdb file as output from whatever > process you used? I would import into swiss pdb to fix the sidechanes and nomenclature. Then I would manually edit out the first phosphorus and its oxygens on both chain a and x. Then I would rename any atoms (ie OP1 - O1P or C* - C'). > > > > > > > > > editconf -bt cubic -f "$base".gro -o "$base".gro -c -d 1.5 > > > > > > > > genbox -cp "$base".gro -cs spc216.gro -o "$base"_b4ion.gro -p "$base".top > > > > grompp -f em.mdp -c "$base"_b4ion.gro -p "$base".top -o "$base"_b4ion.tpr > > > > genion -s "$base"_b4ion.tpr -o "$base"_b4em.gro -neutral -conc 0.001 > > -pname NA -nname CL -g "$base"_ion.log -p "$base".top > > > > > > > > Here I have some issues too, but it is intermittent. I have a charge of > > -563 but genion wants to add 580+ NA and one if any CL giving me a > > possitive charge in the system. My thought is that -neutral would give > > me a neutral system? > > > > > > Unfortunately, genion doesn't deal well with large numbers for some reason. If > you know how many ions you need to neutralize, do it manually with -np or -nn. > > Also of interest: how did you get this net charge? When I processed 1HNX (after > re-building missing atoms, I have a charge in excess of -1000. Sorry for this exact run I was using a truncated pdb to limit atom numbers. It is over 1000 for the whole ribosome subunit. I have been doing it manually when genion is wrong. > > > > > > > > > grompp -f em.mdp -c "$base"_b4em.gro -p "$base".top -o "$base"_em.tpr > > > > > > > > mdrun -v -s "$base"_em.tpr -c "$base"_after_em.gro -g emlog.log > > > > I have set this up to be very thorough and it can run for 1000+ steps. > > My problem is it never gives me a good Fmax. I am always above 1e-4. > > > > An Fmax of 0.0001 is extremely stringent and unlikely to be produced for most > systems in the absence of exhaustive EM using double precision. Such > minimization is usually unnecessary for most systems, unless you're doing normal > modes. The fact that you are "above" 0.0001 is thus unsurprising. How far > above are you? Hard numbers, please, and do remind everyone of what your EM > settings are. Sorry my energies would never get below 1e4 or 1. EM output Potential Energy = -2.4193396e+07 Maximum force = 2.2069838e+04 on atom 18961 Norm of force = 3.3190449e+01 EM Settings: define = -DPOSRES_WATER integrator = steep nsteps = 1 nstcomm = 1 emtol = 200 emstep = 0.0001 nstcgsteep = 10 ns-type = Grid rlist = 0.9 coulombtype = PME rcoulomb-switch = 0 rcoulomb-switch = 0 pme_order = 6 optimize_fft = yes PR output (2 fs steps) No previous checkpoint file present, assuming this is a new run. Getting Loaded... Reading file 1HNX-15_pr.tpr, VERSION 4.5 (single precision) Starting 2 threads Loaded with Money Making 1D domain decomposition 2 x 1 x 1 starting mdrun 'Protein in water' 100 steps, 0.2 ps. step 0 Step 21 Warning: pressure scaling more than 1%, mu: 0.98428 0.98428 0.98428 Step 21 Warning: pressure sc
[gmx-users] Can gromacs developers write code to put an external force on atoms in a specified region in later versions?
Dear gromacs developers, As far as I know, in the latest gromacs version, one only can put an external force on a group of atoms, but can not put external force on atoms in a specific region, for instance, the region from z = 4 nm to z = 6 nm. Some other softwares like NAMD seem to have such function. I wonder if gromacs developers can add this function, assuming it is a easy job, to later versions for it is useful for some special purposes. Best wishes, Zhongqiao Hu Research fellow Dept of Chem Biomol Engineer National University of Singapore -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/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/Support/Mailing_Lists
