Aw: [gmx-users] Free energy calculation: merge the topology of 2 molecules

[lloyd riggs]

I've had 2 problems like this.  1 was solved by doing all eq to a good degree first in one thread, then the domain decomposition worked in 8 or 16...the secound I had to break down the charge groups in the .itp (cg) into smaller charge groups and it worked,
there might be better suggestions though.

 

Stephan

 

Gesendet: Donnerstag, 30. Mai 2013 um 20:48 Uhr
Von: "Dejun Lin" <dejun....@gmail.com>
An: gmx-users@gromacs.org
Betreff: [gmx-users] Free energy calculation: merge the topology of 2 molecules

Hi all,

I'm trying to set-up a free energy calculation where a molecule has +2
charge in it's native state (state A) and no charge in the mutant (state
B). Since the molecule has net +2 charge, I have to add counter-ions to
neutralize the system in state A. But in order to transform it to state B
and still maintain a neutral system, the counter-ions have to be
transformed too. I tried only transforming only the target molecule not the
ions but the simulation crashes very quickly.

I searched the gmx-users archives and found some suggestion about merging
the topology definition of ions into that of the molecule under one
[moleculetype] section. I tried that but mdrun warned me with tons of
"inconsistent shift":

There were 2 inconsistent shifts. Check your topology
There were 18 inconsistent shifts. Check your topology
There were 16 inconsistent shifts. Check your topology
There were 12 inconsistent shifts. Check your topology
There were 16 inconsistent shifts. Check your topology
...

and the simulation can't be run in parallel because mdrun would just quick
and complain about not being able to do domain decomposition:

There is no domain decomposition for 16 nodes that is compatible with the
given box and a minimum cell size of 29.6188 nm

I guess the issue is Gromacs thinks those counter-ions belongs chemically
to the target molecule although they are actually not in close proximity in
space, which mess up the DD. partition.

I wonder if there's a way to get around that.

Thanks,
Dejun
--
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/Support/Mailing_Lists/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/Support/Mailing_Lists

-- 
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/Support/Mailing_Lists/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/Support/Mailing_Lists