[gmx-users] Re: Gromacs 54a7 force field

[Justin Lemkul]

On 7/18/12 5:51 PM, Rajat Desikan wrote:
> "54A7 also introduced changes to the Gromos96 lipid parameters"
> How will this change my inclusion of the berger lipid parameters? Any thing
> that I should pay special attention to? Are there other lipid parameters
> more compatible?

There are better force fields for lipids, but they require the use of CHARMM. 
There may be other suitable united atom force fields.  My comment was 
intentionally generic; I don't know how well 54A7 pairs with the Berger 
parameters.  In theory, it should be fine, but since Gromos96 parameters for 
lipids have been tweaked, maybe the balance of forces in the parameters for 
proteins and lipids will be affected.  There was a new lipid atom type 
introduced to make Gromos96 lipids better.  They're still not as good as CHARMM, 
though.

> I heard from a faculty member at our Institute that the 53a6 is a bad ff for
> a protein with a lot of alpha helices for longer simulations. She apparently
> saw the helices unravel when they were supposed to be stable.
>

This is a hot topic.  Conventional wisdom does demonstrate that 53A6 
destabilizes helices, but others will contend that 53A6 is perfectly suitable 
for reproducing many experimental observables, including NMR signals.  I tend to 
think that there is some bias, but all force fields have shortcomings.  54A7 is 
certainly better in terms of helical stabilization, from what I can see.

-Justin

-- 
========================================

Justin A. Lemkul, Ph.D.
Research Scientist
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================