[gmx-users] Re: opls atom types and charges

[Justin A. Lemkul]

Sebastian Kruggel wrote:
> hi justin,
> 
> thanks for the quick answer and ...
> 
>  > Either way, parameterization is going to be hard.
> 
> so you don't think the use of default parameters of gmx-ff as described 
> in the drug-enzyme-tutorial to be useful (to give reasonable results) if 
> i want to get an impression of the stability of docked poses and for the 
> calculation of binding free energies?
> 
> simulations don't crash the way described in
> http://davapc1.bioch.dundee.ac.uk/prodrg/gmx.pdf
> - i only thought that oplsaa would probably be the 'better choice'?

 From experience I can tell you that PRODRG parameters, at face value, are 
generally unreliable.  The charges and charge groups it provides do not give 
good results when put through the tests that proper parameters should.  This is 
not always the case (depending on the complexity of your molecule), but for the 
compounds I have dealt with, this has definitely been true.  PRODRG is a nice 
tool because it saves you the work of building a topology from scratch, but it 
is still your job to come up with parameters consistent with the original 
derivation of the force field.

As for the "better choice," all force fields have pros and cons.  It's your job 
to read the literature for what's known about each and make an intelligent 
choice.  This is not a trivial decision, and can heavily influence the results 
you get.  Using the ffgmx force field (as in the tutorial) is an especially bad 
idea, since (as pdb2gmx warns you) this force field is ancient and deprecated.

-Justin

> 
> thanks for your estimation,
> sebastian
> 
> 
> 
> 

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

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

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