On Wed, Feb 16, 2011, Bongkeun Kim wrote:
> I did the same thing with ff99SB force field and I got the same error.
> I generated input files from tleap by using the following sequence.
> sequence {NGLU GLU ARG VAL GLN ASP ASN ILE VAL ASN ARG ILE SER ASP ARG CLEU
> }
> It worked well with the single MD with pmemd.cuda.
> sander.mpi with REMD only has this error and adding ACE cap onto the
> first GLU makes me avoid this error.
Can you post what your groupfile and input files look like? It would help
also if you posted the prmtops and input coordinate files. The more work it
is for someone to reproduce the problem, the less likely they are to examine
it. This is an unusual problem, but there might even be simple things wrong
with the REMD inputs that are hard to imagine with the current (lack of)
information, but which might become obvious when someone saw your actual input
files.
And, of course, you may be able to solve this problem yourself: start with the
working REMD examples in the test cases (e.g. $AMBERHOME/test/rem_gb_2rep,
etc.) and try to slowly morph them into your system. For example, use the
same groupfile and input files as in the test case, but substitute in your
prmtop and incprd files. This is all a part of narrowing down the problem.
....dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Feb 17 2011 - 05:30:02 PST
