Hello all,
I am running on a simulation on the microbial photosystem I (pdb: 1JB0)
which is a very, very large system. I have developed parameters for all the
ligands in the system and I am running it in explicit solvent with
neutralizing explicit Na+ ions. The published structure has several
structural clashes and I had to use the XMIN or LMOD minimization approaches
for minimization the starting structure. I have successfully completed two
stages of minimization, heated equilibration, and non-restrained
equilibration for 500 ps. After approximately 4 ns of production MD, I run
into vlimit was exceeded on step ..../vmax errors. Applying very weak
restraints the vlimit errors go away, however I would prefer to have
production MD without the caveat of restraints. Playing with the
restraints, I was able to deduce that the protein sidechains were the main
contributors the vlimit errors. I was unable to visualize the movement that
caused the vlimit errors (despite writing to the trajectory and restart
files for every step). More than likely the movement result from
hydrophobically buried protein sidechains that have build up enough energy
to move out of the bad alignments of the starting structure.
*I need ideas of how minimize these buried protein sidechains*.
Normally I would use swiss pdb viewer and minimize the structure but the
inclusion of non-protein components prevents that approach. I have taken
the structure from the moment that the clashes appear and repeated the
minimization and equilibration but I continue to run into problems down
stream. I have run MD at 10K and then heated it up but I still run into
vlimit errors when the temperatures reach 100K. I would appreciate any
ideas/approaches that others have used with large systems with bad starting
coordinates.
Thank you,
Scott
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed May 18 2011 - 10:00:02 PDT
