[gmx-users] using SETTLE for constraints

[Dmitri Dubov]

> Berk wrote:

> 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!

> 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

> Dmitri Dubov wrote:
> > 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?
 
> 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.

> > 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>