[gmx-users] Lipid simulations in 3.3

[Alan Dodd]

Is anyone still using the old lipid-modified ffgmx
forcefield in v3.3?  If so, have you encountered any
problems with it?
I've been running sensible-looking bilayer+peptide
simulations for over a year now with 3.2.1 and ffgmx. 
Recently we have been trying to compare speed with
3.3, but structure files that work perfectly happily
with old gromacs rapidly crash with the mdrun of new
gromacs.  All the input files we're using are
identical between new and old, with the exception of
ffgmx.hdb which I've had to take from the new version.
 A 3.2.1 .tpr file will run on 3.3 (although it
produces a whole bunch of LINCS errors/rotations of
bonds that don't occur in 3.2.1), suggesting to me
that the problem lies with the interpretation of the
old forcefield.


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com