Re: [gmx-users] Setting up an infinitely hard wall

[Amit Choubey]

Hi Berk,

I am mailing about the issue of setting up infinite potential well.

I now want to simplify the task of inversing the velocity and mirroring the
particle by the following procedure. If a particle is found in z wrote:

>  Hi,
>
> If you have a wall potential, you can reweight the configurations
> with weight 0 if one or more particles are beyond the wall
> and exp(Vwall/kT) if no particles are beyond the wall.
> This will only work efficiently if you choose the wall potential in a smart
> way.
> I managed to get an efficiency of 70%.
>
> If you have constraints, an atom that goes beyond the wall is directly
> coupled to the atoms it is constrained to. You will have to work out
> the equations for such a situation. It might be as simple that you can
> just correct x and v for the atom that went beyond the wall and then
> apply the constraints as normal.
>
> Berk
>
> --
> Date: Fri, 23 Oct 2009 00:26:08 -0700
> Subject: Re: [gmx-users] Setting up an infinitely hard wall
>
> From: kgp.a...@gmail.com
> To: gmx-users@gromacs.org
>
> Hi Berk,
>
> Thank you for the response.
>
> I have obtained distributions with in infinite wall by simulating with an
> softer wall
> and unbiasing with configurations with the wall potential.
>
>
> Could you explain the above ? I am not sure if i get your point.
>
>
> But if you need the dynamics, or you want less hassle during the
> simulation,
> you will have to do a bit more effort in coding an infinitely hard wall.
>
> You will not only have to inverse the velocity, but also mirror the
> position
> of the particle in the wall.
>
>
> That's true. I am sorry i didnt mention that.
>
>
> This should on require a few lines of code in do_update_md
> in src/mdlib/update.c.
>
>
> Great, I will try this as soon as possible.
>
>
> Note that this will only work easily when you do not have constraints
> present.
> With constraints things get much more complicated.
>
>
> I do have constraints present. Thank you for pointing this. I am working
> with SPC water and I should, in principle be able to figure out the
> co-ordinates of all atoms in the molecule if I am given one of the water's
> atom's co-ordinate. Does that sound ok?
>
> Thanks for the input again,
> Amit
>
>
>
> Berk
>
> --
> Date: Thu, 22 Oct 2009 14:34:06 -0700
> From: kgp.a...@gmail.com
> To: gmx-users@gromacs.org
> Subject: [gmx-users] Setting up an infinitely hard wall
>
>
> Hi everyone,
>
> I am sending this email again hoping for any quick input for my question.
>
>
>
> I have been trying to set up an "infinitely" hard potential wall.
>
> I tried to use the available wall options and could not really get it to do
> what i needed. I wanted a steep repulsive potential but when i created that,
> the system was blowing up, reason being that it requires smaller time step
> and i cant afford to have smaller time step.
>
> My idea to overcome this issue is to just reverse the velocity of the
> particle whenever it hits the wall. I am not sure if there is any thing in
> GROMACS which does this but any suggestions will be very helpful.
>
> If there is nothing set up for something like above, i would like to play
> around with the code to figure it out. If this is the case, could somebody
> direct me to a starting point.
>
>  Thank you
> Amit
>
>
> --
> New Windows 7: Find the right PC for you. Learn 
> more.
>
> ___
> 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
>
>
>
> --
> Express yourself instantly with MSN Messenger! MSN 
> Messenger
>
> ___
> 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 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] MVAPICH2 mdrun problem for SD and MD, GROMACS 4.0.5

[Mark Abraham]

Daniel Adriano Silva M wrote:

Mark,

Then I will try with other versions, check gmx-tests and not mix
compilers, thank you much by your help. Last question, I am afraid to
ask, but I have some simulations that were ran with this setup,
energy, temp, and pressure looks good on all, and it ran by pretty
long time (several ns) without any apparent errors, do you think that
there could be any way to check if these md-sims results are actually
wrong?


That's a hard question, because MD with even the same GROMACS version 
with different compilers can diverge rapidly. If you still had the old 
version, you might try a new mdrun -reprod with nstlog=1 on both 
versions. If they give the same output, you might then have some 
confidence that the old version was only wrong under some conditions, 
and hopefully all of those provoked crashes. However it would be likely 
that they wouldn't give the same output even if both were functioning 
correctly. Still, the energy differences over the first few MD steps 
should be down at about the precision reported in the .log file, and 
that's useful knowledge.


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] MVAPICH2 mdrun problem for SD and MD, GROMACS 4.0.5

[Daniel Adriano Silva M]

Mark,

Then I will try with other versions, check gmx-tests and not mix
compilers, thank you much by your help. Last question, I am afraid to
ask, but I have some simulations that were ran with this setup,
energy, temp, and pressure looks good on all, and it ran by pretty
long time (several ns) without any apparent errors, do you think that
there could be any way to check if these md-sims results are actually
wrong?

Thanks
Daniel

2009/10/30 Mark Abraham :
> Daniel Adriano Silva M wrote:
>>
>> Mark,
>>
>> Thank you!!! It was non directly MPI, since I tried the fresh
>> installation with intel and gcc but none worked, then i realized that
>> my compilation of FFTW could be the problem (compiled with icc 11),
>> then I recompiled FFTW with gcc and linking to this new fftw3.2.2 I
>> recompiled gromacs booth with gcc and intel, booth compilations seems
>> to work. Do you think that it could be any problem at mix gcc-fftw and
>> intel mdrun??? I will continue the tests with the new gmx-tests and
>> other TPRs.
>
> There have been known issues with linking and some versions of Intel
> compilers under various conditions with stuff in general and GROMACS in
> particular. I've never cared to understand why. Not mixing compilers is
> usually a good idea if there are problems.
>
> Mark
>
>> 2009/10/30 Mark Abraham :
>>>
>>> Daniel Adriano Silva M wrote:

 Mark,

 I will test, but please tell me: do you think MPI linking problem
 could lead to problems with some dynamics and not with others as
 happens to me? and also note that all my test where made with mvapich2
 (even that with one core). Please justin, what do you think about?
>>>
>>> For example, observing symptoms from buffer overruns can be sensitive to
>>> the
>>> actual calculation being run because it can depend how the actual memory
>>> gets laid out and used. So here, that might translate to the kind of
>>> system
>>> being simulated, and the number of cores used. Such an overrun might be
>>> present in the code all the time, or only exists after a linking
>>> mismatch,
>>> or similar.
>>>
>>> Mark
>>>
 Thanks
 Daniel

 2009/10/30 Mark Abraham :
>
> Daniel Adriano Silva M wrote:
>>
>> Dear Gromacs users,
>>
>> I am experimenting the next problem on an infiniband-cluster (8
>> intel-cores per node, GROMACS compiled with icc 11.1, all run through
>> mvapich2):
>>
>> I have a molecule (protein 498aa, solvated or in vacuum I get the same
>> problem at any box shape), when I try to SD minimize it with
>> mvapich2-mdrun, it minimizes well with 1, 2 or 3 cores and reaches
>> convergence in around 1000steps, however any further combination (4,5,
>> ...n cores) makes it to immediately stop (less than 20 steps) with:
>>
>> "Steepest Descents converged to machine precision...".
>>
>> Further if I take "the 1 core minimized structure" and try to make a
>> solvated-pr dynamics(2fs, MD, NTP, etc.) it also works with 1
>> processor, but with more cores it begins immediately to bring LINCS
>> warnings: and dies:
>>
>> "Too many LINCS warnings" or "Water molecule starting at atom 16221
>> can not be settled"
>>
>> For a "long time" I had made another md simulations on this cluster
>> with the same mdps and other proteic systems, and I only see this
>> behavior with this particular protein, of course  before send this
>> mail I re-tested previuos-working tprs.
>> Finally, the most suspicious is that I have another very similar
>> 8-core box (with the same processors) but with gromacs gcc compiled,
>> and it actually runs very well the same problematic molecule (even the
>> same tpr) with mpi and 8-cores.
>> What do you think??? Please, if you have some tpr to test something
>> send
>> it.
>
> I'd guess you're having some problem with (dynamic) linking of the MPI
> library. Perhaps the version of some library has changed since
> recently,
> etc. I'd suggest compiling two fresh copies of GROMACS with either icc
> and
> gcc on the troublesome machine and seeing what happens with them.
>
> Mark
> ___
> 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
>
 ___
 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. Us

Re: [gmx-users] MVAPICH2 mdrun problem for SD and MD, GROMACS 4.0.5

[Mark Abraham]

Daniel Adriano Silva M wrote:

Mark,

Thank you!!! It was non directly MPI, since I tried the fresh
installation with intel and gcc but none worked, then i realized that
my compilation of FFTW could be the problem (compiled with icc 11),
then I recompiled FFTW with gcc and linking to this new fftw3.2.2 I
recompiled gromacs booth with gcc and intel, booth compilations seems
to work. Do you think that it could be any problem at mix gcc-fftw and
intel mdrun??? I will continue the tests with the new gmx-tests and
other TPRs.


There have been known issues with linking and some versions of Intel 
compilers under various conditions with stuff in general and GROMACS in 
particular. I've never cared to understand why. Not mixing compilers is 
usually a good idea if there are problems.


Mark


2009/10/30 Mark Abraham :

Daniel Adriano Silva M wrote:

Mark,

I will test, but please tell me: do you think MPI linking problem
could lead to problems with some dynamics and not with others as
happens to me? and also note that all my test where made with mvapich2
(even that with one core). Please justin, what do you think about?

For example, observing symptoms from buffer overruns can be sensitive to the
actual calculation being run because it can depend how the actual memory
gets laid out and used. So here, that might translate to the kind of system
being simulated, and the number of cores used. Such an overrun might be
present in the code all the time, or only exists after a linking mismatch,
or similar.

Mark


Thanks
Daniel

2009/10/30 Mark Abraham :

Daniel Adriano Silva M wrote:

Dear Gromacs users,

I am experimenting the next problem on an infiniband-cluster (8
intel-cores per node, GROMACS compiled with icc 11.1, all run through
mvapich2):

I have a molecule (protein 498aa, solvated or in vacuum I get the same
problem at any box shape), when I try to SD minimize it with
mvapich2-mdrun, it minimizes well with 1, 2 or 3 cores and reaches
convergence in around 1000steps, however any further combination (4,5,
...n cores) makes it to immediately stop (less than 20 steps) with:

"Steepest Descents converged to machine precision...".

Further if I take "the 1 core minimized structure" and try to make a
solvated-pr dynamics(2fs, MD, NTP, etc.) it also works with 1
processor, but with more cores it begins immediately to bring LINCS
warnings: and dies:

"Too many LINCS warnings" or "Water molecule starting at atom 16221
can not be settled"

For a "long time" I had made another md simulations on this cluster
with the same mdps and other proteic systems, and I only see this
behavior with this particular protein, of course  before send this
mail I re-tested previuos-working tprs.
Finally, the most suspicious is that I have another very similar
8-core box (with the same processors) but with gromacs gcc compiled,
and it actually runs very well the same problematic molecule (even the
same tpr) with mpi and 8-cores.
What do you think??? Please, if you have some tpr to test something send
it.

I'd guess you're having some problem with (dynamic) linking of the MPI
library. Perhaps the version of some library has changed since recently,
etc. I'd suggest compiling two fresh copies of GROMACS with either icc
and
gcc on the troublesome machine and seeing what happens with them.

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 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 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 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] Re: NAN in g_anaeig -proj

[Tsjerk Wassenaar]

Hi Alex,

Unfortunately I don't have time to dig into that now. It might indeed
be a bug though and you might want to consider file a bugzilla report:
http://bugzilla.gromacs.org
That will put it on the table of the developers.

Cheers,

Tsjerk


2009/10/30 alexander yakovenko :
> Hi Tsjerk!
> That is where I encountered NANs problem (except it was 10 eigenvectos, but
> for 1 NANs appears as well). I used -over option to check eigenvectors only.
>
> Now it seems for me that it is a bug and not my stupid typing mistake. What
> can really help in this case (IMHO!) is my problem in your debugger - I
> found that the problem appears even if I try to project trajectory of single
> frame gro file:
> g_anaeig_d -v sss_1000_eigenvec.trr -f sss_after_mdsi_1000.gro -s
> sss_mdsi_1000.tpr -first 1 -last 1 -n sss.ndx -proj
> so I can send you gzipped problem.
> Regards,
> Alex.
>
>
>
> Message: 6
> Date: Fri, 30 Oct 2009 12:53:28 +0100
> From: Tsjerk Wassenaar 
> Subject: Re: [gmx-users] NAN in g_anaeig -proj
> To: Discussion list for GROMACS users 
> Message-ID:
> <8ff898150910300453n550a6358sf596cff4771c9...@mail.gmail.com>
> Content-Type: text/plain; charset=ISO-8859-1
>
> Hi Alex,
>
> Sorry for not replying earlier. Still haven't had time to do checks
> myself. But have you already tried to see what happens if you only
> request projections?:
>
> g_anaeig_d -v sss_1000_eigenvec.trr -f sss_mdsi_1000.trr -s
> sss_mdsi_1000.tpr -first 1 -last 1 -n sss.ndx -proj
>
> Cheers,
>
> Tsjerk
>
> 2009/10/29 alexander yakovenko :
>> Hi all!
>> I just wonder if anyone run (know solution) into a problem trying to
>> project
>> a trajectory on the eigenvector(s) (with g_anaeig -proj ) from covariace
>> matrix.
>> All projections I have calculated are nan. However, the eigenvalues are OK
>> and the g_anaeig -comp -v2 -eig2 -over options works OK so eigenvectors
>> seems are OK  (but g_anaeig -2d fails to nan too). I am using
>> gromacs-4.0.5
>> on x86-64 CentOS5 (compiled with gcc-34).
>> Regards,
>> Alex.
>> P.S. The sequence of commands that reveals the problem is attached bellow:
>>
>> g_covar_d -f sss_mdsi_1000.trr -s sss_mdsi_1000.tpr -n sss.ndx -o
>> sss_1000_eigenval.xvg -xpm sss_1000_covar.xpm -v sss_1000_eigenvec.trr
>> -mwa
>> -l
>> ...
>> Option Filename  Type Description
>> 
>>   -f sss_mdsi_1000.trr  Input    Trajectory: xtc trr trj gro g96 pdb
>> cpt
>>   -s sss_mdsi_1000.tpr  Input    Structure+mass(db): tpr tpb tpa gro
>> g96
>>    pdb
>>   -n    sss.ndx  Input, Opt!  Index file
>>   -o sss_1000_eigenval.xvg  Output   xvgr/xmgr file
>>   -v sss_1000_eigenvec.trr  Output   Full precision trajectory: trr
>> trj
>>    cpt
>>  -av    average.pdb  Output   Structure file: gro g96 pdb
>>   -l  covar.log  Output   Log file
>> -ascii    covar.dat  Output, Opt. Generic data file
>> -xpm sss_1000_covar.xpm  Output, Opt! X PixMap compatible matrix file
>> -xpma    covara.xpm  Output, Opt. X PixMap compatible matrix file
>>
>> Option   Type   Value   Description
>> --
>> -[no]h   bool   no  Print help info and quit
>> -nice    int    19  Set the nicelevel
>> -b   time   0   First frame (ps) to read from trajectory
>> -e   time   0   Last frame (ps) to read from trajectory
>> -dt  time   0   Only use frame when t MOD dt = first time (ps)
>> -tu  enum   ps  Time unit: ps, fs, ns, us, ms or s
>> -[no]xvgr    bool   yes Add specific codes (legends etc.) in the
>> output
>>     xvg files for the xmgrace program
>> -[no]fit bool   yes Fit to a reference structure
>> -[no]ref bool   no  Use the deviation from the conformation in the
>>     structure file instead of from the average
>> -[no]mwa bool   yes Mass-weighted covariance analysis
>> -last    int    -1  Last eigenvector to write away (-1 is till the
>>     last)
>> -[no]pbc bool   yes Apply corrections for periodic boundary
>> conditions
>>
>> Reading file sss_mdsi_1000.tpr, VERSION 4.0.5 (double precision)
>> Reading file sss_mdsi_1000.tpr, VERSION 4.0.5 (double precision)
>>
>> Choose a group for the least squares fit
>> Group 0 (  System) has 31748 elements
>> ...
>> Group    20 ( active_site) has    54 elements
>> Select a group: 20
>> Selected 20: 'active_site'
>>
>> Choose a group for the covariance analysis
>> Group 0 (  System) has 31748 elements
>> ...
>> Group    20 ( active_site) has    54 elements
>> Select a group: 20
>> Selected 20: 'active_site'
>> Calculating the average structure ...
>> trn version: GMX_trn_file (double precision)
>> Last frame   2000 time 2000.000
>>
>> Constructing covariance matrix (162x162) ...
>> Last frame   2000 time

Re: [gmx-users] MVAPICH2 mdrun problem for SD and MD, GROMACS 4.0.5

[Daniel Adriano Silva M]

Mark,

Thank you!!! It was non directly MPI, since I tried the fresh
installation with intel and gcc but none worked, then i realized that
my compilation of FFTW could be the problem (compiled with icc 11),
then I recompiled FFTW with gcc and linking to this new fftw3.2.2 I
recompiled gromacs booth with gcc and intel, booth compilations seems
to work. Do you think that it could be any problem at mix gcc-fftw and
intel mdrun??? I will continue the tests with the new gmx-tests and
other TPRs.

Thanks
Daniel

2009/10/30 Mark Abraham :
> Daniel Adriano Silva M wrote:
>>
>> Mark,
>>
>> I will test, but please tell me: do you think MPI linking problem
>> could lead to problems with some dynamics and not with others as
>> happens to me? and also note that all my test where made with mvapich2
>> (even that with one core). Please justin, what do you think about?
>
> For example, observing symptoms from buffer overruns can be sensitive to the
> actual calculation being run because it can depend how the actual memory
> gets laid out and used. So here, that might translate to the kind of system
> being simulated, and the number of cores used. Such an overrun might be
> present in the code all the time, or only exists after a linking mismatch,
> or similar.
>
> Mark
>
>> Thanks
>> Daniel
>>
>> 2009/10/30 Mark Abraham :
>>>
>>> Daniel Adriano Silva M wrote:

 Dear Gromacs users,

 I am experimenting the next problem on an infiniband-cluster (8
 intel-cores per node, GROMACS compiled with icc 11.1, all run through
 mvapich2):

 I have a molecule (protein 498aa, solvated or in vacuum I get the same
 problem at any box shape), when I try to SD minimize it with
 mvapich2-mdrun, it minimizes well with 1, 2 or 3 cores and reaches
 convergence in around 1000steps, however any further combination (4,5,
 ...n cores) makes it to immediately stop (less than 20 steps) with:

 "Steepest Descents converged to machine precision...".

 Further if I take "the 1 core minimized structure" and try to make a
 solvated-pr dynamics(2fs, MD, NTP, etc.) it also works with 1
 processor, but with more cores it begins immediately to bring LINCS
 warnings: and dies:

 "Too many LINCS warnings" or "Water molecule starting at atom 16221
 can not be settled"

 For a "long time" I had made another md simulations on this cluster
 with the same mdps and other proteic systems, and I only see this
 behavior with this particular protein, of course  before send this
 mail I re-tested previuos-working tprs.
 Finally, the most suspicious is that I have another very similar
 8-core box (with the same processors) but with gromacs gcc compiled,
 and it actually runs very well the same problematic molecule (even the
 same tpr) with mpi and 8-cores.
 What do you think??? Please, if you have some tpr to test something send
 it.
>>>
>>> I'd guess you're having some problem with (dynamic) linking of the MPI
>>> library. Perhaps the version of some library has changed since recently,
>>> etc. I'd suggest compiling two fresh copies of GROMACS with either icc
>>> and
>>> gcc on the troublesome machine and seeing what happens with them.
>>>
>>> Mark
>>> ___
>>> 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
>>>
>> ___
>> 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
>>
> ___
> 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
>
___
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] Re: NAN in g_anaeig -proj

[alexander yakovenko]

Hi Tsjerk!  
That is where I encountered NANs problem (except it was 10 eigenvectos, but for 
1 NANs appears as well). I used -over option to check eigenvectors only.  
  
Now it seems for me that it is a bug and not my stupid typing mistake. What can 
really help in this case (IMHO!) is my problem in your debugger - I found that 
the problem appears even if I try to project trajectory of single frame gro 
file:  
g_anaeig_d -v sss_1000_eigenvec.trr -f sss_after_mdsi_1000.gro -s 
sss_mdsi_1000.tpr -first 1 -last 1 -n sss.ndx -proj  
so I can send you gzipped problem.  
Regards,  
Alex.  
  

  
Message: 6  
Date: Fri, 30 Oct 2009 12:53:28 +0100  
From: Tsjerk Wassenaar   
Subject: Re: [gmx-users] NAN in g_anaeig -proj  
To: Discussion list for GROMACS users   
Message-ID:  
<8ff898150910300453n550a6358sf596cff4771c9...@mail.gmail.com>  
Content-Type: text/plain; charset=ISO-8859-1  
  
Hi Alex,  
  
Sorry for not replying earlier. Still haven't had time to do checks  
myself. But have you already tried to see what happens if you only  
request projections?:  
  
g_anaeig_d -v sss_1000_eigenvec.trr -f sss_mdsi_1000.trr -s  
sss_mdsi_1000.tpr -first 1 -last 1 -n sss.ndx -proj  
  
Cheers,  
  
Tsjerk  
  
2009/10/29 alexander yakovenko :  
> Hi all!  
> I just wonder if anyone run (know solution) into a problem trying to project  
> a trajectory on the eigenvector(s) (with g_anaeig -proj ) from covariace  
> matrix.  
> All projections I have calculated are nan. However, the eigenvalues are OK  
> and the g_anaeig -comp -v2 -eig2 -over options works OK so eigenvectors  
> seems are OK  (but g_anaeig -2d fails to nan too). I am using gromacs-4.0.5  
> on x86-64 CentOS5 (compiled with gcc-34).  
> Regards,  
> Alex.  
> P.S. The sequence of commands that reveals the problem is attached bellow:  
>  
> g_covar_d -f sss_mdsi_1000.trr -s sss_mdsi_1000.tpr -n sss.ndx -o  
> sss_1000_eigenval.xvg -xpm sss_1000_covar.xpm -v sss_1000_eigenvec.trr -mwa  
> -l  
> ...  
> Option Filename  Type Description  
>   
>   -f sss_mdsi_1000.trr  Input    Trajectory: xtc trr trj gro g96 pdb cpt  
>   -s sss_mdsi_1000.tpr  Input    Structure+mass(db): tpr tpb tpa gro g96  
>    pdb  
>   -n    sss.ndx  Input, Opt!  Index file  
>   -o sss_1000_eigenval.xvg  Output   xvgr/xmgr file  
>   -v sss_1000_eigenvec.trr  Output   Full precision trajectory: trr trj  
>    cpt  
>  -av    average.pdb  Output   Structure file: gro g96 pdb  
>   -l  covar.log  Output   Log file  
> -ascii    covar.dat  Output, Opt. Generic data file  
> -xpm sss_1000_covar.xpm  Output, Opt! X PixMap compatible matrix file  
> -xpma    covara.xpm  Output, Opt. X PixMap compatible matrix file  
>  
> Option   Type   Value   Description  
> --  
> -[no]h   bool   no  Print help info and quit  
> -nice    int    19  Set the nicelevel  
> -b   time   0   First frame (ps) to read from trajectory  
> -e   time   0   Last frame (ps) to read from trajectory  
> -dt  time   0   Only use frame when t MOD dt = first time (ps)  
> -tu  enum   ps  Time unit: ps, fs, ns, us, ms or s  
> -[no]xvgr    bool   yes Add specific codes (legends etc.) in the output  
>     xvg files for the xmgrace program  
> -[no]fit bool   yes Fit to a reference structure  
> -[no]ref bool   no  Use the deviation from the conformation in the  
>     structure file instead of from the average  
> -[no]mwa bool   yes Mass-weighted covariance analysis  
> -last    int    -1  Last eigenvector to write away (-1 is till the  
>     last)  
> -[no]pbc bool   yes Apply corrections for periodic boundary  
> conditions  
>  
> Reading file sss_mdsi_1000.tpr, VERSION 4.0.5 (double precision)  
> Reading file sss_mdsi_1000.tpr, VERSION 4.0.5 (double precision)  
>  
> Choose a group for the least squares fit  
> Group 0 (  System) has 31748 elements  
> ...  
> Group    20 ( active_site) has    54 elements  
> Select a group: 20  
> Selected 20: 'active_site'  
>  
> Choose a group for the covariance analysis  
> Group 0 (  System) has 31748 elements  
> ...  
> Group    20 ( active_site) has    54 elements  
> Select a group: 20  
> Selected 20: 'active_site'  
> Calculating the average structure ...  
> trn version: GMX_trn_file (double precision)  
> Last frame   2000 time 2000.000  
>  
> Constructing covariance matrix (162x162) ...  
> Last frame   2000 time 2000.000  
> Read 2001 frames  
>  
> Trace of the covariance matrix: 6.22756 (u nm^2)  
>  
> 100%  
> Diagonalizing ...  
>  
> Sum of the eigenvalues: 6.22756 (u nm^2)  
>  
> Writing eigenvalues to sss_1000_eigenval.xvg

Re: [gmx-users] MVAPICH2 mdrun problem for SD and MD, GROMACS 4.0.5

[Mark Abraham]

Daniel Adriano Silva M wrote:

Mark,

I will test, but please tell me: do you think MPI linking problem
could lead to problems with some dynamics and not with others as
happens to me? and also note that all my test where made with mvapich2
(even that with one core). Please justin, what do you think about?


For example, observing symptoms from buffer overruns can be sensitive to 
the actual calculation being run because it can depend how the actual 
memory gets laid out and used. So here, that might translate to the kind 
of system being simulated, and the number of cores used. Such an overrun 
might be present in the code all the time, or only exists after a 
linking mismatch, or similar.


Mark


Thanks
Daniel

2009/10/30 Mark Abraham :

Daniel Adriano Silva M wrote:

Dear Gromacs users,

I am experimenting the next problem on an infiniband-cluster (8
intel-cores per node, GROMACS compiled with icc 11.1, all run through
mvapich2):

I have a molecule (protein 498aa, solvated or in vacuum I get the same
problem at any box shape), when I try to SD minimize it with
mvapich2-mdrun, it minimizes well with 1, 2 or 3 cores and reaches
convergence in around 1000steps, however any further combination (4,5,
...n cores) makes it to immediately stop (less than 20 steps) with:

"Steepest Descents converged to machine precision...".

Further if I take "the 1 core minimized structure" and try to make a
solvated-pr dynamics(2fs, MD, NTP, etc.) it also works with 1
processor, but with more cores it begins immediately to bring LINCS
warnings: and dies:

"Too many LINCS warnings" or "Water molecule starting at atom 16221
can not be settled"

For a "long time" I had made another md simulations on this cluster
with the same mdps and other proteic systems, and I only see this
behavior with this particular protein, of course  before send this
mail I re-tested previuos-working tprs.
Finally, the most suspicious is that I have another very similar
8-core box (with the same processors) but with gromacs gcc compiled,
and it actually runs very well the same problematic molecule (even the
same tpr) with mpi and 8-cores.
What do you think??? Please, if you have some tpr to test something send
it.

I'd guess you're having some problem with (dynamic) linking of the MPI
library. Perhaps the version of some library has changed since recently,
etc. I'd suggest compiling two fresh copies of GROMACS with either icc and
gcc on the troublesome machine and seeing what happens with them.

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 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 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] Problem with CGMD

[Justin A. Lemkul]

Anirban Ghosh wrote:

Hello Justin,

Thanx for the reply
I tried with dt=0.02 and also with dt=0.01. but still its crashing. It 
is crashing in the very first step. So nothings there in trr file.
Are the options given in the MDP file correct? Actually my EM ran quite 
well. Please advice.




If your system is collapsing at the very first step, I would argue that EM was 
not complete.  What potential energy and maximum force did you achieve?


If the .mdp file came from the MARTINI group, I would think it is appropriate; I 
have only done limited testing and found that I needed to change dt to get a 
stable simulation.


-Justin

--


Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
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

Re: [gmx-users] Problem with CGMD

[Anirban Ghosh]

Hello Justin,

Thanx for the reply
I tried with dt=0.02 and also with dt=0.01. but still its crashing. It is
crashing in the very first step. So nothings there in trr file.
Are the options given in the MDP file correct? Actually my EM ran quite
well. Please advice.

Regards,

Anirban

On Fri, Oct 30, 2009 at 7:28 PM, Justin A. Lemkul  wrote:

>
>
> Anirban Ghosh wrote:
>
>> Hi ALL,
>>
>> Thanks Justin for your reply.
>> Now after solvating my CG system of protein in lipid bilayer, I did EM and
>> the final energy values were reasonable. But while running MD I am getting
>> the following error:
>>
>>
>> ---
>> Fatal error:
>> Number of grid cells is zero. Probably the system and box collapsed
>>
>> ---
>>
>> In the archives I found this error and possible solution "to check the
>> structure". But that is for all-atom system and according to  me my CG
>> system is OK. But I am not sure about the mdp parameters that I got from a
>> example script from MARTINI site for running lipid MD. So can you please
>> check the mdp file and suggest a solution for this problem.
>>
>
> That advice is not specific for atomistic MD.  If there is something
> inherently wrong with your structure, then you will have problems.  I'm
> curious to know how you decided that everything was fine, because it
> certainly is not!
>
> Have you watched the trajectory?  Does the crash happen immediately?  In my
> experience, a dt > 0.02 always crashes when using MARTINI, but I know the
> authors claim it can go as high as 0.04.
>
> -Justin
>
> ---
>> ; STANDARD MD INPUT OPTIONS FOR MARTINI 2.0
>> ;
>> ; for use with GROMACS 3.3
>> ;
>>
>> ; VARIOUS PREPROCESSING OPTIONS =
>> title= Martini
>> cpp  = /usr/bin/cpp
>>
>> ; RUN CONTROL PARAMETERS =
>> ; MARTINI - Most simulations are stable with dt=40 fs,
>> ; some (especially rings) require 20-30 fs.
>> ; The range of time steps used for parametrization ; is 20-40 fs, using
>> smaller time steps is therefore not recommended.
>>
>> integrator   = md
>> ; start time and timestep in ps
>> tinit= 0.0
>> dt   = 0.030
>> nsteps   = 90
>> ; number of steps for center of mass motion removal =
>> nstcomm  = 1
>> comm-grps =
>>
>> ; OUTPUT CONTROL OPTIONS =
>> ; Output frequency for coords (x), velocities (v) and forces (f) =
>> nstxout  = 5000
>> nstvout  = 5000
>> nstfout  = 0
>> ; Output frequency for energies to log file and energy file =
>> nstlog   = 1000
>> nstenergy= 1000
>> ; Output frequency and precision for xtc file =
>> nstxtcout= 1000
>> xtc_precision= 100
>> ; This selects the subset of atoms for the xtc file. You can =
>> ; select multiple groups. By default all atoms will be written. =
>> xtc-grps =
>> ; Selection of energy groups =
>> energygrps   =
>>
>> ; NEIGHBORSEARCHING PARAMETERS =
>> ; MARTINI - no need for more frequent updates
>> ; or larger neighborlist cut-off due
>> ; to the use of shifted potential energy functions.
>>
>> ; nblist update frequency =
>> nstlist  = 10
>> ; ns algorithm (simple or grid) =
>> ns_type  = grid
>> ; Periodic boundary conditions: xyz or none =
>> pbc  = xyz
>> ; nblist cut-off =
>> rlist= 1.2
>>
>> ; OPTIONS FOR ELECTROSTATICS AND VDW =
>> ; MARTINI - vdw and electrostatic interactions are used
>> ; in their shifted forms. Changing to other types of
>> ; electrostatics will affect the general performance of
>> ; the model.
>>
>> ; Method for doing electrostatics =
>> coulombtype  = Shift
>> rcoulomb_switch  = 0.0
>> rcoulomb = 1.2
>> ; Dielectric constant (DC) for cut-off or DC of reaction field =
>> epsilon_r= 15
>> ; Method for doing Van der Waals =
>> vdw_type = Shift
>> ; cut-off lengths=
>> rvdw_switch  = 0.9
>> rvdw = 1.2
>> ; Apply long range dispersion corrections for Energy and Pressure =
>> DispCorr = No
>>
>> ; OPTIONS FOR WEAK COUPLING ALGORITHMS =
>> ; MARTINI - normal temperature and pressure coupling schemes
>> ; can be used. It is recommended to couple individual groups
>> ; in your system seperately.
>>
>> ; Temperature coupling   =
>> tcoupl   = Berendsen
>> ; Groups to couple separately =
>> tc-grps  = Protein DSPC W
>> ; Time constant (ps) and reference temperature (K) =
>> tau_t=

Re: [gmx-users] Problem with CGMD

[Justin A. Lemkul]

Anirban Ghosh wrote:

Hi ALL,

Thanks Justin for your reply.
Now after solvating my CG system of protein in lipid bilayer, I did EM 
and the final energy values were reasonable. But while running MD I am 
getting the following error:


---
Fatal error:
Number of grid cells is zero. Probably the system and box collapsed
---

In the archives I found this error and possible solution "to check the 
structure". But that is for all-atom system and according to  me my CG 
system is OK. But I am not sure about the mdp parameters that I got from 
a example script from MARTINI site for running lipid MD. So can you 
please check the mdp file and suggest a solution for this problem.


That advice is not specific for atomistic MD.  If there is something inherently 
wrong with your structure, then you will have problems.  I'm curious to know how 
you decided that everything was fine, because it certainly is not!


Have you watched the trajectory?  Does the crash happen immediately?  In my 
experience, a dt > 0.02 always crashes when using MARTINI, but I know the 
authors claim it can go as high as 0.04.


-Justin


---
; STANDARD MD INPUT OPTIONS FOR MARTINI 2.0
;
; for use with GROMACS 3.3
;

; VARIOUS PREPROCESSING OPTIONS =
title= Martini
cpp  = /usr/bin/cpp

; RUN CONTROL PARAMETERS =
; MARTINI - Most simulations are stable with dt=40 fs,
; some (especially rings) require 20-30 fs.
; The range of time steps used for parametrization 
; is 20-40 fs, using smaller time steps is therefore not recommended.


integrator   = md
; start time and timestep in ps
tinit= 0.0
dt   = 0.030
nsteps   = 90
; number of steps for center of mass motion removal =
nstcomm  = 1
comm-grps =

; OUTPUT CONTROL OPTIONS =
; Output frequency for coords (x), velocities (v) and forces (f) =
nstxout  = 5000
nstvout  = 5000
nstfout  = 0
; Output frequency for energies to log file and energy file =
nstlog   = 1000
nstenergy= 1000
; Output frequency and precision for xtc file =
nstxtcout= 1000
xtc_precision= 100
; This selects the subset of atoms for the xtc file. You can =
; select multiple groups. By default all atoms will be written. =
xtc-grps =
; Selection of energy groups =
energygrps   =

; NEIGHBORSEARCHING PARAMETERS =
; MARTINI - no need for more frequent updates
; or larger neighborlist cut-off due
; to the use of shifted potential energy functions.

; nblist update frequency =
nstlist  = 10
; ns algorithm (simple or grid) =
ns_type  = grid
; Periodic boundary conditions: xyz or none =
pbc  = xyz
; nblist cut-off =
rlist= 1.2

; OPTIONS FOR ELECTROSTATICS AND VDW =
; MARTINI - vdw and electrostatic interactions are used
; in their shifted forms. Changing to other types of
; electrostatics will affect the general performance of
; the model.

; Method for doing electrostatics =
coulombtype  = Shift
rcoulomb_switch  = 0.0
rcoulomb = 1.2
; Dielectric constant (DC) for cut-off or DC of reaction field =
epsilon_r= 15
; Method for doing Van der Waals =
vdw_type = Shift
; cut-off lengths=
rvdw_switch  = 0.9
rvdw = 1.2
; Apply long range dispersion corrections for Energy and Pressure =
DispCorr = No

; OPTIONS FOR WEAK COUPLING ALGORITHMS =
; MARTINI - normal temperature and pressure coupling schemes
; can be used. It is recommended to couple individual groups
; in your system seperately.

; Temperature coupling   =
tcoupl   = Berendsen
; Groups to couple separately =
tc-grps  = Protein DSPC W
; Time constant (ps) and reference temperature (K) =
tau_t= 0.3 0.3 0.3
ref_t= 315 315 315
; Pressure coupling  =
Pcoupl   = berendsen
Pcoupltype   = isotropic
; Time constant (ps), compressibility (1/bar) and reference P (bar) =
tau_p= 3.0
compressibility  = 3e-5
ref_p= 1.0

; GENERATE VELOCITIES FOR STARTUP RUN =
gen_vel  = no
gen_temp = 315
gen_seed = 666

; OPTIONS FOR BONDS =
; MARTINI - for ring systems constraints are defined
; which are best handled using Lincs.

constraints  = none
; Type of constraint algorithm =
constraint_algorithm 

[gmx-users] Problem with CGMD

[Anirban Ghosh]

Hi ALL,

Thanks Justin for your reply.
Now after solvating my CG system of protein in lipid bilayer, I did EM and
the final energy values were reasonable. But while running MD I am getting
the following error:

---
Fatal error:
Number of grid cells is zero. Probably the system and box collapsed
---

In the archives I found this error and possible solution "to check the
structure". But that is for all-atom system and according to  me my CG
system is OK. But I am not sure about the mdp parameters that I got from a
example script from MARTINI site for running lipid MD. So can you please
check the mdp file and suggest a solution for this problem.
---
; STANDARD MD INPUT OPTIONS FOR MARTINI 2.0
;
; for use with GROMACS 3.3
;

; VARIOUS PREPROCESSING OPTIONS =
title= Martini
cpp  = /usr/bin/cpp

; RUN CONTROL PARAMETERS =
; MARTINI - Most simulations are stable with dt=40 fs,
; some (especially rings) require 20-30 fs.
; The range of time steps used for parametrization
; is 20-40 fs, using smaller time steps is therefore not recommended.

integrator   = md
; start time and timestep in ps
tinit= 0.0
dt   = 0.030
nsteps   = 90
; number of steps for center of mass motion removal =
nstcomm  = 1
comm-grps =

; OUTPUT CONTROL OPTIONS =
; Output frequency for coords (x), velocities (v) and forces (f) =
nstxout  = 5000
nstvout  = 5000
nstfout  = 0
; Output frequency for energies to log file and energy file =
nstlog   = 1000
nstenergy= 1000
; Output frequency and precision for xtc file =
nstxtcout= 1000
xtc_precision= 100
; This selects the subset of atoms for the xtc file. You can =
; select multiple groups. By default all atoms will be written. =
xtc-grps =
; Selection of energy groups =
energygrps   =

; NEIGHBORSEARCHING PARAMETERS =
; MARTINI - no need for more frequent updates
; or larger neighborlist cut-off due
; to the use of shifted potential energy functions.

; nblist update frequency =
nstlist  = 10
; ns algorithm (simple or grid) =
ns_type  = grid
; Periodic boundary conditions: xyz or none =
pbc  = xyz
; nblist cut-off =
rlist= 1.2

; OPTIONS FOR ELECTROSTATICS AND VDW =
; MARTINI - vdw and electrostatic interactions are used
; in their shifted forms. Changing to other types of
; electrostatics will affect the general performance of
; the model.

; Method for doing electrostatics =
coulombtype  = Shift
rcoulomb_switch  = 0.0
rcoulomb = 1.2
; Dielectric constant (DC) for cut-off or DC of reaction field =
epsilon_r= 15
; Method for doing Van der Waals =
vdw_type = Shift
; cut-off lengths=
rvdw_switch  = 0.9
rvdw = 1.2
; Apply long range dispersion corrections for Energy and Pressure =
DispCorr = No

; OPTIONS FOR WEAK COUPLING ALGORITHMS =
; MARTINI - normal temperature and pressure coupling schemes
; can be used. It is recommended to couple individual groups
; in your system seperately.

; Temperature coupling   =
tcoupl   = Berendsen
; Groups to couple separately =
tc-grps  = Protein DSPC W
; Time constant (ps) and reference temperature (K) =
tau_t= 0.3 0.3 0.3
ref_t= 315 315 315
; Pressure coupling  =
Pcoupl   = berendsen
Pcoupltype   = isotropic
; Time constant (ps), compressibility (1/bar) and reference P (bar) =
tau_p= 3.0
compressibility  = 3e-5
ref_p= 1.0

; GENERATE VELOCITIES FOR STARTUP RUN =
gen_vel  = no
gen_temp = 315
gen_seed = 666

; OPTIONS FOR BONDS =
; MARTINI - for ring systems constraints are defined
; which are best handled using Lincs.

constraints  = none
; Type of constraint algorithm =
constraint_algorithm = Lincs
; Do not constrain the start configuration =
unconstrained_start  = no
; Highest order in the expansion of the constraint coupling matrix =
lincs_order  = 4
; Lincs will write a warning to the stderr if in one step a bond =
; rotates over more degrees than =
lincs_warnangle  = 30

Re: [gmx-users] user defined potential function for 1-4 interaction

[M Hafizur Rahman]

Hi Berk:

Thanks for your eply. I guess I would supply two column in the table for the 
bond type 9 [r(nm) and W(kJ/mol) ]. In that case (according to the user manual) 
what is the significance of k? how  k will come into the calculation? If I 
select "constraints = all bond" will that affect bond type 9?

Thanks,
M. H. Rahman
  - Original Message - 
  From: Berk Hess 
  To: Discussion list for GROMACS users 
  Sent: Friday, October 09, 2009 10:02 AM
  Subject: RE: [gmx-users] user defined potential function for 1-4 interaction


  Hi,

  Unfortunately this is not possible.
  If you do not have to many pairs, you can change all (or all minus one type)
  from pairs to tabulated bonds (bonds type 9) in your topology.
  Then you can supply as many different bond tables as you like.
  Note that this will supply a single potential though,
  not separate Coulomb, dispersion and repulsion tables.

  Berk


--
  From: mhrah...@dal.ca
  To: gmx-users@gromacs.org
  Subject: Re: [gmx-users] user defined potential function for 1-4 interaction
  Date: Fri, 9 Oct 2009 09:48:19 -0300


  I think there must be a way in gromacs to pass tabulated potenttial for 1-4 
interaction for two groups as interaction between different pairs can be 
significantly different and can change the structure of an indiviual molecule. 
  I would request:  Please let me know if it is possible to pass tabulated 
potential for pair inteaction for two groups. if my method, which you will see 
below, is wrong please let me know the correct way. I searched previous 
communications regarding tabulated potential but could not get my answer. I 
hope I am not wasing your time.

  Thanks.


  - Original Message - 
From: LuLanyuan 
To: gmx-users@gromacs.org 
Sent: Thursday, October 08, 2009 8:04 PM
Subject: RE: [gmx-users] user defined potential function for 1-4 interaction


Hi,
As I remember, you can only use one table for 1-4 interactions. Don't know 
if it's changed
in the latest version.
Lanyuan

> From: mhrah...@dal.ca
> To: gmx-users@gromacs.org
> Subject: Re: [gmx-users] user defined potential function for 1-4 
interaction
> Date: Thu, 8 Oct 2009 19:10:39 -0300
> 
> Hi all:
> I posted this earlier but did not get any help I really need to slove 
this 
> issue.
> 
> What should be the way to pass a formatted table for the 1-4 interaction
> for two or more different groups? Reading page 150 and 180 of manual 4.0, 
> I
> understand user defined potential functions can be passed for many groups
> using energygrp_table in .mdp file but nothing is clearly mentioned
> regarding pair interaction
> 
> I have two grps for 1-4 interactions MTN and SFT.
> mdrun command includes:
> -table table.xvg -tablep pair.xvg
> ; .mdp file contains,
> energygrps = MTN SFT
> energygrp_table: MTN MTN MTN SFT
> 
> I passed three tables for pair interaction in the
> name pair.xvg, pair_MTN_MTN.xvg and pair_MTN_SFT.xvg.
> I see the later two tabels are not recognized. If pair_MTN_MTN.xvg and 
> pair_MTN_SFT.xvg are not supplied GROMACS donot issue any warning.
> I am afraid that I failed to submit separate table for pair interaction 
> properly in this way.
> 
> Thanks,
> M.H.Rahman 
> 
> ___
> 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/ma iling_lists/users.php



搜索本应是快乐的，不是么? 快乐搜索，有问必应！微软隆重推出! 立即查看！ 



___
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


--
  Express yourself instantly with MSN Messenger! MSN Messenger 


--


  ___
  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 (

Re: [gmx-users] Solvating a CG system

[Justin A. Lemkul]

Anirban Ghosh wrote:

Hi ALL,

I have a CG protein in a CG lipid bilayer. I have packed the lipids 
around the protein using InflateGRO, following Justin's tutorial. Now I 
want to solvate it with CG waters. In all-atom we need to set a higher 
value for VanderWaal's radii for C in vdwradii.dat file. What should it 
be like for solvating my CG system, since there are no "C" here? Is 
there any different file for that? I tried using genbox normally and 
fornd CG waters inserted between the CG protein beads. How to solve this?

Any suggestion is welcome.



Add entries in vdwradii.dat for the atom types in your CG model, or remove the 
stray particles manually or with a script like those found here:


http://www.gromacs.org/Documentation/How-tos/Membrane_Simulations

-Justin


Regards,

Anirban




___
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


--


Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
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

Re: [gmx-users] MVAPICH2 mdrun problem for SD and MD, GROMACS 4.0.5

[Daniel Adriano Silva M]

Mark,

I will test, but please tell me: do you think MPI linking problem
could lead to problems with some dynamics and not with others as
happens to me? and also note that all my test where made with mvapich2
(even that with one core). Please justin, what do you think about?

Thanks
Daniel

2009/10/30 Mark Abraham :
> Daniel Adriano Silva M wrote:
>>
>> Dear Gromacs users,
>>
>> I am experimenting the next problem on an infiniband-cluster (8
>> intel-cores per node, GROMACS compiled with icc 11.1, all run through
>> mvapich2):
>>
>> I have a molecule (protein 498aa, solvated or in vacuum I get the same
>> problem at any box shape), when I try to SD minimize it with
>> mvapich2-mdrun, it minimizes well with 1, 2 or 3 cores and reaches
>> convergence in around 1000steps, however any further combination (4,5,
>> ...n cores) makes it to immediately stop (less than 20 steps) with:
>>
>> "Steepest Descents converged to machine precision...".
>>
>> Further if I take "the 1 core minimized structure" and try to make a
>> solvated-pr dynamics(2fs, MD, NTP, etc.) it also works with 1
>> processor, but with more cores it begins immediately to bring LINCS
>> warnings: and dies:
>>
>> "Too many LINCS warnings" or "Water molecule starting at atom 16221
>> can not be settled"
>>
>> For a "long time" I had made another md simulations on this cluster
>> with the same mdps and other proteic systems, and I only see this
>> behavior with this particular protein, of course  before send this
>> mail I re-tested previuos-working tprs.
>> Finally, the most suspicious is that I have another very similar
>> 8-core box (with the same processors) but with gromacs gcc compiled,
>> and it actually runs very well the same problematic molecule (even the
>> same tpr) with mpi and 8-cores.
>> What do you think??? Please, if you have some tpr to test something send
>> it.
>
> I'd guess you're having some problem with (dynamic) linking of the MPI
> library. Perhaps the version of some library has changed since recently,
> etc. I'd suggest compiling two fresh copies of GROMACS with either icc and
> gcc on the troublesome machine and seeing what happens with them.
>
> Mark
> ___
> 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
>
___
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] Shear viscosity

[Vitaly V. Chaban]

Hi all,

I am trying to master the art of shear viscosity calculation using
"g_energy -vis". Plotting the first column in visco.xvg over time, the
nearly straight line is obtained. Is it OK? How can one get the
viscosity value from these data? My simulation details for this task
are: NVT, 298K, shifted LJ with a big Rcut, PME, the substance is an
ionic liquid, simulation time is 2 ns with 1 fs timestep and every
10fs saving to ener.edr.

I suspect something is definitely wrong in my calculation of
viscosity. I tried the same procedure with SPC water (NPT, 1ns, PME,
shifted LJ) and got the complicated curve (in visco.xvg, first column)
with an average value of 0.55 cP. This result seems to be quite more
realistic as for me. Then what is the possible source of problem in my
above setup?

Thanks for any suggestions.
Vitaly
___
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] How to calculate the lifetime of one hydrogen bond

[Justin A. Lemkul]

Dechang Li wrote:
Dear all, 

	I want to calculate the lifetime of one hydrogen bond. I used g_hbond to do this. Here is the out put: 


   Hydrogen bond thermodynamics at T = 298.15 K
--
Type  Rate (1/ps)  Time (ps)  DG (kJ/mol)
Forward-0.001   -1750.289-666.000
Backward   -0.001   -930.928-666.000
One-way 0.000   6622.058  26.338
Integral0.001739.899  20.905
Relaxation  0.778  1.286   5.152

	Which is the lifetime of the hydrogen bond? What does the "Forward", "Backward" etc. mean? 


You might find this thread interesting:

http://lists.gromacs.org/pipermail/gmx-users/2009-February/039955.html

-Justin




Best regards,
2009-1030


= 
Dechang Li, Ph.D Candidate

Department of Engineering Mechanics
Tsinghua University
Beijing 100084
P.R. China 

Tel:   +86-10-62773574(O) 
Email: lid...@mails.tsinghua.edu.cn

=

  


___
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



--


Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
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] How to calculate the lifetime of one hydrogen bond

[Dechang Li]

Dear all, 

I want to calculate the lifetime of one hydrogen bond. I used g_hbond 
to do this. Here is the out put: 

   Hydrogen bond thermodynamics at T = 298.15 K
--
Type  Rate (1/ps)  Time (ps)  DG (kJ/mol)
Forward-0.001   -1750.289-666.000
Backward   -0.001   -930.928-666.000
One-way 0.000   6622.058  26.338
Integral0.001739.899  20.905
Relaxation  0.778  1.286   5.152

Which is the lifetime of the hydrogen bond? What does the "Forward", 
"Backward" etc. mean? 


Best regards,
2009-1030


= 
Dechang Li, Ph.D Candidate
Department of Engineering Mechanics
Tsinghua University
Beijing 100084
P.R. China 

Tel:   +86-10-62773574(O) 
Email: lid...@mails.tsinghua.edu.cn
=

  

___
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] Solvating a CG system

[Anirban Ghosh]

Hi ALL,

I have a CG protein in a CG lipid bilayer. I have packed the lipids around
the protein using InflateGRO, following Justin's tutorial. Now I want to
solvate it with CG waters. In all-atom we need to set a higher value for
VanderWaal's radii for C in vdwradii.dat file. What should it be like for
solvating my CG system, since there are no "C" here? Is there any different
file for that? I tried using genbox normally and fornd CG waters inserted
between the CG protein beads. How to solve this?
Any suggestion is welcome.

Regards,

Anirban
___
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] NAN in g_anaeig -proj

[Tsjerk Wassenaar]

Hi Alex,

Sorry for not replying earlier. Still haven't had time to do checks
myself. But have you already tried to see what happens if you only
request projections?:

g_anaeig_d -v sss_1000_eigenvec.trr -f sss_mdsi_1000.trr -s
sss_mdsi_1000.tpr -first 1 -last 1 -n sss.ndx -proj

Cheers,

Tsjerk

2009/10/29 alexander yakovenko :
> Hi all!
> I just wonder if anyone run (know solution) into a problem trying to project
> a trajectory on the eigenvector(s) (with g_anaeig -proj ) from covariace
> matrix.
> All projections I have calculated are nan. However, the eigenvalues are OK
> and the g_anaeig -comp -v2 -eig2 -over options works OK so eigenvectors
> seems are OK  (but g_anaeig -2d fails to nan too). I am using gromacs-4.0.5
> on x86-64 CentOS5 (compiled with gcc-34).
> Regards,
> Alex.
> P.S. The sequence of commands that reveals the problem is attached bellow:
>
> g_covar_d -f sss_mdsi_1000.trr -s sss_mdsi_1000.tpr -n sss.ndx -o
> sss_1000_eigenval.xvg -xpm sss_1000_covar.xpm -v sss_1000_eigenvec.trr -mwa
> -l
> ...
> Option Filename  Type Description
> 
>   -f sss_mdsi_1000.trr  Input    Trajectory: xtc trr trj gro g96 pdb cpt
>   -s sss_mdsi_1000.tpr  Input    Structure+mass(db): tpr tpb tpa gro g96
>    pdb
>   -n    sss.ndx  Input, Opt!  Index file
>   -o sss_1000_eigenval.xvg  Output   xvgr/xmgr file
>   -v sss_1000_eigenvec.trr  Output   Full precision trajectory: trr trj
>    cpt
>  -av    average.pdb  Output   Structure file: gro g96 pdb
>   -l  covar.log  Output   Log file
> -ascii    covar.dat  Output, Opt. Generic data file
> -xpm sss_1000_covar.xpm  Output, Opt! X PixMap compatible matrix file
> -xpma    covara.xpm  Output, Opt. X PixMap compatible matrix file
>
> Option   Type   Value   Description
> --
> -[no]h   bool   no  Print help info and quit
> -nice    int    19  Set the nicelevel
> -b   time   0   First frame (ps) to read from trajectory
> -e   time   0   Last frame (ps) to read from trajectory
> -dt  time   0   Only use frame when t MOD dt = first time (ps)
> -tu  enum   ps  Time unit: ps, fs, ns, us, ms or s
> -[no]xvgr    bool   yes Add specific codes (legends etc.) in the output
>     xvg files for the xmgrace program
> -[no]fit bool   yes Fit to a reference structure
> -[no]ref bool   no  Use the deviation from the conformation in the
>     structure file instead of from the average
> -[no]mwa bool   yes Mass-weighted covariance analysis
> -last    int    -1  Last eigenvector to write away (-1 is till the
>     last)
> -[no]pbc bool   yes Apply corrections for periodic boundary
> conditions
>
> Reading file sss_mdsi_1000.tpr, VERSION 4.0.5 (double precision)
> Reading file sss_mdsi_1000.tpr, VERSION 4.0.5 (double precision)
>
> Choose a group for the least squares fit
> Group 0 (  System) has 31748 elements
> ...
> Group    20 ( active_site) has    54 elements
> Select a group: 20
> Selected 20: 'active_site'
>
> Choose a group for the covariance analysis
> Group 0 (  System) has 31748 elements
> ...
> Group    20 ( active_site) has    54 elements
> Select a group: 20
> Selected 20: 'active_site'
> Calculating the average structure ...
> trn version: GMX_trn_file (double precision)
> Last frame   2000 time 2000.000
>
> Constructing covariance matrix (162x162) ...
> Last frame   2000 time 2000.000
> Read 2001 frames
>
> Trace of the covariance matrix: 6.22756 (u nm^2)
>
> 100%
> Diagonalizing ...
>
> Sum of the eigenvalues: 6.22756 (u nm^2)
>
> Writing eigenvalues to sss_1000_eigenval.xvg
>
> Writing reference, average structure & eigenvectors 1--162 to
> sss_1000_eigenvec.trr
>
> Wrote the log to covar.log
>
> gcq#86: "Shake Barrels Of Whisky Down My Throat" (Throwing Muses)
>
>
>
> g_anaeig_d -v sss_1000_eigenvec.trr -eig sss_1000_eigenval.xvg -v2
> sss_1000_eigenvec.trr -eig2 sss_1000_eigenval.xvg -f sss_mdsi_1000.trr -s
> sss_mdsi_1000.tpr -first 1 -last 10 -n sss.ndx -proj -over
>
> ...
>
> Option Filename  Type Description
> 
>   -v sss_1000_eigenvec.trr  Input    Full precision trajectory: trr trj
>    cpt
>  -v2 sss_1000_eigenvec.trr  Input, Opt!  Full precision trajectory: trr trj
>    cpt
>   -f sss_mdsi_1000.trr  Input, Opt!  Trajectory: xtc trr trj gro g96 pdb cpt
>   -s sss_mdsi_1000.tpr  Input, Opt!  Structure+mass(db): tpr tpb tpa gro g96
>    pdb
>   -n    sss.ndx  Input, Opt!  Index file
> -eig sss_1000_eigenval.xvg  Input, Opt!  xvgr/xmgr file
> -eig2sss_1000_eigenval.x

Re: [gmx-users] Constraints and ligands

[Mark Abraham]

P.R.Anand Narayanan wrote:

Dear users,
I just have 2 queries
1) Is it safe to use "united atoms" constraint for positional 
restraining and md simulation of an enzyme docked to the substrate? if 
yes, what is the KEYWORD to be typed in the mdp file and in what way is 
it simpler than the "all-bonds" constraint?


Constraints and restraints are distinct concepts in GROMACS. United 
atoms are something else again. Please do some background reading in the 
manual and ask a focussed question then :-)


2) How can a chemical ligand be bonded to a protein ie; with what type 
of bond formation and how are the residues in the protein chosen for the 
bonding to be done with the ligand?


It can be bonded whatever way makes sense for your investigation, but 
producing a corresponding topology need not be easy, even if suitable 
force field parameters already exist. You will need a thorough knowledge 
of chapter 5 of the manual. As an exercise that will teach you the 
necessary skills, design a .top file by hand for a molecule that is the 
ligand bound to just the relevant residue, ignoring the rest of the protein.


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] MVAPICH2 mdrun problem for SD and MD, GROMACS 4.0.5

[Mark Abraham]

Daniel Adriano Silva M wrote:

Dear Gromacs users,

I am experimenting the next problem on an infiniband-cluster (8
intel-cores per node, GROMACS compiled with icc 11.1, all run through
mvapich2):

I have a molecule (protein 498aa, solvated or in vacuum I get the same
problem at any box shape), when I try to SD minimize it with
mvapich2-mdrun, it minimizes well with 1, 2 or 3 cores and reaches
convergence in around 1000steps, however any further combination (4,5,
...n cores) makes it to immediately stop (less than 20 steps) with:

"Steepest Descents converged to machine precision...".

Further if I take "the 1 core minimized structure" and try to make a
solvated-pr dynamics(2fs, MD, NTP, etc.) it also works with 1
processor, but with more cores it begins immediately to bring LINCS
warnings: and dies:

"Too many LINCS warnings" or "Water molecule starting at atom 16221
can not be settled"

For a "long time" I had made another md simulations on this cluster
with the same mdps and other proteic systems, and I only see this
behavior with this particular protein, of course  before send this
mail I re-tested previuos-working tprs.
Finally, the most suspicious is that I have another very similar
8-core box (with the same processors) but with gromacs gcc compiled,
and it actually runs very well the same problematic molecule (even the
same tpr) with mpi and 8-cores.
What do you think??? Please, if you have some tpr to test something send it.


I'd guess you're having some problem with (dynamic) linking of the MPI 
library. Perhaps the version of some library has changed since recently, 
etc. I'd suggest compiling two fresh copies of GROMACS with either icc 
and gcc on the troublesome machine and seeing what happens with them.


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] Number of solvent molecules

[Mark Abraham]

P.R.Anand Narayanan wrote:

Dear users,

How is it possible to fix the number of solvent molecules and the size 
of the box used so that I can maintain exactly the same environment for 
different runs.


What are you planning to vary?

For eg, if I have 84000 water solvent molecules and the dimensions of 
6.7 6.73 4.71 nm; how can I use the same system for a different run?? 


If you actually mean to ask this question, then you provide the 
coordinate and .top files of "the system" as input to another invocation 
of grompp in the normal way. I think you've not described your situation 
fully, however!


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] Re: Number of solvent molecules

[Vitaly V. Chaban]

> Dear users,
>
> How is it possible to fix the number of solvent molecules and the size of the 
> box used so that I can maintain exactly the same environment for different 
> runs.
> For eg, if I have 84000 water solvent molecules and the dimensions of 6.7 
> 6.73 4.71 nm; how can I use the same system for a

Your solvent molecules are in conf(out).gro, they do not appear or
disappear during MD. To keep the box size constant, use NVT (no
barostat in md. mdp).

Vitaly
___
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] Different temperatures for different groups, even with Nose-Hoover

[XAvier Periole]

The comment from Justin and your answer to it were pretty convincing.
I think the problem is solved :)).

Tom is correct to get T with g_energy on a simulation that ran with one
Tcoup, you need to rerun the trajectory (trr) with a tpr in which you
define two Tcoup. This should give you T for each group.

XAvier.


On Oct 29, 2009, at 11:11 PM, Michael Lerner wrote:




On Thu, Oct 29, 2009 at 5:31 PM, XAvier Periole   
wrote:


On Oct 29, 2009, at 9:26 PM, Michael Lerner wrote:


I calculated temperatures with

g_traj -f md2.trr -s md2.tpr -n index.ndx -ot -ng 5
I did not know you could get the temperature throught g_traj ... the  
-ot option is the one

giving you the temperature?


Yes.

Do you get the same temperature difference with g_energy? Or may be  
the info is not
in there? May be you could rerun you trajectory using a new tpr  
where you define two
different temp coupling groups. And see if you observe the same  
temperature given by

g_energy.

It's possible that I'm doing something wrong, but I couldn't  
convince g_energy to give me the temperature of different groups  
when I used a single thermostat for the entire system. If I use two  
temperature coupling groups, I get the expected results.


As Justin pointed out, it looks like the error was mine for not  
understanding the details of g_traj. When I manually correct for the  
degrees of freedom, I get temperatures for Protein/Sol that are  
close enough to each other (i.e. within a couple of standard  
deviations).


Thanks,

-Michael


Thanks,

-Michael

-- begin md2.mdp --

; RUN CONTROL PARAMETERS =
integrator   = md
; start time and timestep in ps =
tinit= 0.0
dt   = 0.020
nsteps   = 2550
; mode for center of mass motion removal
comm-mode= linear
; number of steps for center of mass motion removal
nstcomm  = 1
; group(s) for center of mass motion removal
comm-grps= System


; OUTPUT CONTROL OPTIONS =
; Output frequency for coords (x), velocities (v) and forces (f) =
nstxout  = 5
nstvout  = 5
nstfout  = 0
; Output frequency for energies to log file and energy file =
nstlog   = 2500
nstenergy= 2500
; Output frequency and precision for xtc file =
nstxtcout= 2500
xtc_precision= 100
; This selects the subset of atoms for the xtc file. You can =
; select multiple groups. By default all atoms will be written. =
xtc-grps =
; Selection of energy groups =
energygrps   = System Protein W WF

; NEIGHBORSEARCHING PARAMETERS =
; nblist update frequency =
nstlist  = 5
; ns algorithm (simple or grid) =
ns_type  = grid
; Periodic boundary conditions: xyz or none =
pbc  = xyz
; nblist cut-off =
rlist= 1.2

; OPTIONS FOR ELECTROSTATICS AND VDW =
; Method for doing electrostatics =
coulombtype  = Shift
rcoulomb_switch  = 0.0
rcoulomb = 1.2
; Dielectric constant (DC) for cut-off or DC of reaction field =
epsilon_r= 15
; Method for doing Van der Waals =
vdw_type = Shift
; cut-off lengths=
rvdw_switch  = 0.9
rvdw = 1.2
; Apply long range dispersion corrections for Energy and Pressure =
DispCorr = No
; Spacing for the PME/PPPM FFT grid =
fourierspacing   = 0.12
; FFT grid size, when a value is 0 fourierspacing will be used =
fourier_nx   = 10
fourier_ny   = 10
fourier_nz   = 10
; EWALD/PME/PPPM parameters =
pme_order= 4
ewald_rtol   = 1e-05
epsilon_surface  = 0
optimize_fft = no

; OPTIONS FOR WEAK COUPLING ALGORITHMS =
; Temperature coupling   =
tcoupl   = Nose-Hoover
; Groups to couple separately =
tc-grps  = System
; Time constant (ps) and reference temperature (K) =
tau_t= 2
ref_t= 323
; Pressure coupling  =
Pcoupl   = Parrinello-Rahman
Pcoupltype   = isotropic
; Time constant (ps), compressibility (1/bar) and reference P (bar) =
tau_p= 5
compressibility  = 5e-5
ref_p= 1.0

; SIMULATED ANNEALING CONTROL =
annealing= no

; GENERATE VELOCITIES FOR STARTUP RUN =
continuation  = yes
gen_vel  = no
;gen_temp = 323
;gen_seed = 473529

; OPTIONS FOR BONDS =
constraints  = none
; Type of constraint algorithm =
constraint_algorithm = Lincs
; Relative tolerance of shake =
shake_tol= 0.0001
; Highest order in the expansion of the constraint coupling matrix =
lincs_order  = 4
; Lincs will write a warning to the stderr if in one step a bond =
;