[gmx-users] using SETTLE for constraints

[Dmitri Dubov]

Dear GMX'ers,

I am confused with some basic questions about using SETTLE constraining algorithm in gromacs.

1) This solution of SHAKE is used, particularly, for spc water. So spc.itp file contains the [ settles ] section, and there is no another [ constraints ] within it. But the parameter "constraint-algorithm" in /water/grompp.mdp (see the gromacs native tutorial for spc water simulation) is set to "Lincs". What is the REAL treatment of water constraints in this case?

2) From manual the run parameter "constraint-algorithm" has only two (LINCS or SHAKE) alternatives. How can I specify the SETTLE using?

Regards,
Dmitri
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20090907/a7d07470/attachment.html>

[Mark Abraham]

Post by Dmitri Dubov
Dear GMX'ers,
I am confused with some basic questions about using SETTLE constraining algorithm in gromacs.
1) This solution of SHAKE is used, particularly, for spc water. So spc.itp file contains the [ settles ] section, and there is no another [ constraints ] within it. But the parameter "constraint-algorithm" in /water/grompp.mdp (see the gromacs native tutorial for spc water simulation) is set to "Lincs". What is the REAL treatment of water constraints in this case?

GROMACS transparently uses the best constraint algorithm that you allow
it use on a given molecule. The only way you can *not* use constraints
on water with the standard water .itp files is to #define FLEXIBLE, as
you will see in those .itp files. Since settle is specified there it is
used.

The choice of LINCS or SHAKE pertains only to molecules for which there
is no other constraint option. You can verify this by doing a quick
simulation of pure water specifying either of the above and seeing that
the results are identical. IIRC the .log file makes a note of what
constraints are being used.

Post by Dmitri Dubov
2) From manual the run parameter "constraint-algorithm" has only two (LINCS or SHAKE) alternatives. How can I specify the SETTLE using?

See above. Looking for SETTLE in the manual index would have helped :-)

Mark

[Berk Hess]

Things are a bit confusing here.

SETTLE versus other constraint algorithm is chosen at the topology level
by putting [ settle ] or [ constraints ] sections in your topology.
With of the other constraint algorithms (LINCS or SHAKE) is used
for the constraints in the [ constraints ] sections is set in the mdp file.

For future Gromacs versions I would like to have the [ settle ] sections
in the water topology replaced with [ constraints ] and let mdrun
decide if to use settle.

From a user point of view, this will not change much though,

since settle is only useful for water and a user (nearly) never
writes a topology for water.

Berk

Date: Mon, 7 Sep 2009 18:28:05 +1000
From: Mark.Abraham at anu.edu.au
To: gmx-users at gromacs.org
Subject: Re: [gmx-users] using SETTLE for constraints

Post by Dmitri Dubov
Dear GMX'ers,
I am confused with some basic questions about using SETTLE constraining algorithm in gromacs.
1) This solution of SHAKE is used, particularly, for spc water. So spc.itp file contains the [ settles ] section, and there is no another [ constraints ] within it. But the parameter "constraint-algorithm" in /water/grompp.mdp (see the gromacs native tutorial for spc water simulation) is set to "Lincs". What is the REAL treatment of water constraints in this case?

GROMACS transparently uses the best constraint algorithm that you allow
it use on a given molecule. The only way you can *not* use constraints
on water with the standard water .itp files is to #define FLEXIBLE, as
you will see in those .itp files. Since settle is specified there it is
used.
The choice of LINCS or SHAKE pertains only to molecules for which there
is no other constraint option. You can verify this by doing a quick
simulation of pure water specifying either of the above and seeing that
the results are identical. IIRC the .log file makes a note of what
constraints are being used.

Post by Dmitri Dubov
2) From manual the run parameter "constraint-algorithm" has only two (LINCS or SHAKE) alternatives. How can I specify the SETTLE using?

See above. Looking for SETTLE in the manual index would have helped :-)
Mark
_______________________________________________
gmx-users mailing list gmx-users at 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-request at gromacs.org.
Can't post? Read http://www.gromacs.org/mailing_lists/users.php

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE!
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20090907/041d3696/attachment.html>

[Dmitri Dubov]

Post by Berk Hess
Things are a bit confusing here.
SETTLE versus other constraint algorithm is chosen at the topology level
by putting [ settle ] or [ constraints ] sections in your topology.
With of the other constraint algorithms (LINCS or SHAKE) is used
for the constraints in the [ constraints ] sections is set in the mdp file.

:yes: That is just the thing I need. Thank you!

Post by Berk Hess
For future Gromacs versions I would like to have the [ settle ] sections
in the water topology replaced with [ constraints ] and let mdrun
decide if to use settle.
From a user point of view, this will not change much though,
since settle is only useful for water and a user (nearly) never
writes a topology for water.
Berk
Date: Mon, 7 Sep 2009 18:28:05 +1000
From: Mark.Abraham at anu.edu.au
To: gmx-users at gromacs.org
Subject: Re: [gmx-users] using SETTLE for constraints

Post by Dmitri Dubov
Dear GMX'ers,
I am confused with some basic questions about using SETTLE constraining algorithm in gromacs.
1) This solution of SHAKE is used, particularly, for spc water. So spc.itp file contains the [ settles ] section, and there is no another [ constraints ] within it. But the parameter "constraint-algorithm" in /water/grompp.mdp (see the gromacs native tutorial for spc water simulation) is set to "Lincs". What is the REAL treatment of water constraints in this case?

GROMACS transparently uses the best constraint algorithm that you allow
it use on a given molecule. The only way you can *not* use constraints
on water with the standard water .itp files is to #define FLEXIBLE, as
you will see in those .itp files. Since settle is specified there it is
used.

As I understand, the defining [ constraints ] instead of [ settles ] in topology file gives the way to use both SHAKE and LINCS for water. If you need, of course :-)
The last (not least) question. Whether the SETTLE is always the best algorithm for water? Or: Supposed I would get different results using SETTLE and LINCS. Does it mean that the reason is in bad LINCS adjustment, and the run with LINCS should be rejected or refined?

Post by Berk Hess
The choice of LINCS or SHAKE pertains only to molecules for which there
is no other constraint option. You can verify this by doing a quick
simulation of pure water specifying either of the above and seeing that
the results are identical. IIRC the .log file makes a note of what
constraints are being used.

Post by Dmitri Dubov
2) From manual the run parameter "constraint-algorithm" has only two (LINCS or SHAKE) alternatives. How can I specify the SETTLE using?

See above. Looking for SETTLE in the manual index would have helped :-)
Mark
_______________________________________________
gmx-users mailing list gmx-users at 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-request at gromacs.org.
Can't post? Read http://www.gromacs.org/mailing_lists/users.php

Dmitri
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20090907/a617b74f/attachment.html>