feenstra at chem.vu.nl
Wed Oct 1 16:41:02 CEST 2003
Dallas Warren wrote:
>> You can still first do it 'manually' with trjconv, with the added
>> that it only has to be done once, and you can check if it is done the way
>> you expect it to. Having said that, the last time I tried either -pbc
>> or the 'cluster' option, it didn't work the way I expected (but I also
>> not have time then to get to the bottom of it).
> Yeah, I have some trajectories where it does an OK job, others where I
> have tried so many options and combinations to try and get it to get
> things all together in the center of the box, to no avail.
> What I was actually suggesting is that the scripts can go over the pbc,
> out one side and in the other, as if it was within the box. Since it is
> continuous anyway with the simulation, shouldn't it also be with the
> analysis scripts?
Yes, the code at least is all there. But there are special issues to consider.
For calculating distances, having all *atoms* in the box is easiest. For fitting
and RMSD calculation, however, you'd like whole *molecules*, unless you want
to fit aggregates or other types of grouped molecules (co-factors, ligands),
then you need some way to tell the analysis tool which things belong together.
I think, if we'd use PBC aware routines throughout, all should work itself out
automatically. But, that would need a major overhaul of all tools. Having done
several minor overhauls myself, I know this is no small task...
| | |
| _ _ ___,| K. Anton Feenstra |
| / \ / \'| | | Dept. of Pharmacochem. - Vrije Universiteit Amsterdam |
|( | )| | | De Boelelaan 1083 - 1081 HV Amsterdam - Netherlands |
| \_/ \_/ | | | Tel: +31 20 44 47608 - Fax: +31 20 44 47610 |
| | Feenstra at chem.vu.nl - www.chem.vu.nl/~feenstra/ |
| | "If You See Me Getting High, Knock Me Down" |
| | (Red Hot Chili Peppers) |
More information about the gromacs.org_gmx-users