Hi,
Vlad is right, the length does not matter in this case.
Our force field is indeed based on CHARMM to a rather large extent, but all the charges have been recomputed using RESP and all parameters (expect for covalent bonds and angles) for the for the lipid tails have been derived from scratch.
Regarding the compatibility, we also computed free energies of solvation in water and cyclohexane. These values are close to the experimental ones and the errors in free energies of transfer was essentially due to the errors in the free energy of hydration. Since the carbons and hydrogens of the methylene groups all have a zero charge only LJ interactions are present which means that essentially any AMBER force field could be used (since the LJ parameters has not changed since AMBER99 if I am correct).
The helix simulations show that the interactions between the amino acids and the lipids are not too strong, which can be the case sometimes and the membrane does not shrink.
Obviously the best thing would be to perform more extensive testing of the compatibility but that would be very time consuming. There is not that much experimental data on membrane proteins either...
Vlad, it would have been very interesting to discuss with you during the meeting at Longholmen but unfortunately I could not attend.
Will try to port this force field to AMBER format, but from what I understand it is not possible to specify explicit parameters for 1-4 LJ pairs in a straightforward manner?
Thanks.
Best,
Joakim
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Jun 22 2012 - 04:30:02 PDT
