Le 07/11/2024 à 16:25, Enrico Martinez via AMBER a écrit :
> P.S. just testing the autoimage moveanchor I noticed that it produced a
> very chaotic lipid distribution outside the simulation box, which contained
> only protein and solvent .. ;-D
Hi all,
According to my tests, autoimage do not easily consider lipid like
proteins, or,
probably, lipids move too fast to ensure that the system stays whole.
On one of my systems, I did also found that the protein was moving close
to an edge of the box,
so i understand why the algorithm fails in finding a goog "central point".
Results of my experiments are:
a) for newer systems use a bigger membrane setup with roughly z
direction being half of the xy dimensions,
for instance --dist_wat 50 --distxy_fix 100 in packmol-membgen
b) use "iwrap=1" for the md trajectory
c) for already produced trajectory, my best findings were to:
(a) first use the "unwrap" command to reposition everything in one
box :
unwrap ":1-218" # where :1-218 is the protein, to ensure at least the
protein is in the same box
center ":1-218" mass origin # Center the system on the protein
image origin center familiar # Center everything on the protein
center ':PA,PC,OL' # Then center on the membrane to see evolution of the
system over time
This could not be applied blindly for all systems but allowed me to
produce nice trajectories
of my membrane systems suitable for publication
(
https://www.biorxiv.org/content/10.1101/2024.10.02.616265v1)
HTH,
Stéphane
PS : this problem is complex, I also encounter this problem using GROMACS,
I'm not sure there is a final algorithm able to handle all cases properly...
There are literature concerns about it, though, see for instance:
Bart M. H. Bruininks, Tsjerk A. Wassenaar, and Ilpo Vattulainen
Journal of Chemical Information and Modeling 2023 63 (11), 3448-3452
DOI: 10.1021/acs.jcim.2c01574
--
Maître de conférences Hors Classe, PEDR, HDR
US2B, Nantes Université, CNRS, UMR 6286, Team Structural Bioinformatics, F-44000 Nantes
http://www.us2b.univ-nantes.fr/ -http://www.steletch.org
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Nov 08 2024 - 01:00:03 PST
