[gmx-users] topology treatment in free energy calculations - possible bug

[bharat v. adkar]

On Fri, 4 Jan 2008, Maik Goette wrote:

> Hi
>
> I just found a strange behaviour of GROMACS, when processing topologies with 
> B-values. Maybe I just think of it as unintuitive/bug, but here we go:
>
> e.g. OPLS
>
> Consider an angle given (all atoms have B-values, which angle-parameter can 
> be found by GROMACS in the bonded.itp for the A- and B-state):
>
> Original topology entry:
> 10    12    14     1
>
> Manually edited entry (with the correct ff-term):
> 10    12    14     1     109.700     669.440
>
> Now, a dump from the tpr-files yields the following:
> Original:
> functype[109]=ANGLES, thA= 1.09700e+02, ctA= 6.69440e+02, thB= 1.09500e+02, 
> ctB= 2.92880e+02
>
> Edited:
> functype[109]=ANGLES, thA= 1.09700e+02, ctA= 6.69440e+02, thB= 1.09700e+02, 
> ctB= 6.69440e+02
>
> Now, what obviously happens is, that GROMACS searches the entries of the 
> angle in the FF for both states and uses them in the case of the original 
> topology.
> If one puts a manual entry into the topology for the A-state though, the 
> B-state is simply copied, instead of searched by GROMACS.
>
> This, IMHO, is quite inconvenient. Is this a bug or a feature?

it cannot be called as a bug... when you mention parameters for A state 
in topology explicitly then grompp will expect you to mention the B state 
parameters also, and if it doesn't find it, it will assign A state 
parameter to B state. that is quite well documented, i think, in manual. 
So there is no reason to think it as a bug.

bharat

>
> Regards
>
>

-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.