[Fwd: [gmx-users] preferred/best force fields]

[Mark Abraham]

iulek at interponta.com.br wrote:
> Hi,
> 
>     I see there are many interesting topics going on, maybe due to this 
> I did not receive any answer/suggestion/comment to my question or maybe 
> it is rather naive. Anyway, it would like to listen to the experts if 
> there are no other suggestions to add to what I did below.
>     Thanks,
> 
> J.
> 
> -------- Original Message --------
> Subject:     [gmx-users] preferred/best force fields
> Date:     Fri, 14 Aug 2009 12:35:18 -0300
> From:     iulek at interponta.com.br
> Reply-To:     Discussion list for GROMACS users <gmx-users at gromacs.org>
> To:     gmx-users at gromacs.org
> 
> 
> Dear All,
> 
>     I am new to molecular dynamics, had some tutorials and could run
> some initial tests successfully in that they finished without errors
> (now to analyze if results make sense).
>     Now my challenge is to make a molecular dynamics of a homohexamer
> which binds to a ssDNS and a co-factor, and that loses some of its
> binding capacities when a mutation is present. I could build an
> (non-hydrogen) homology model for the system, but I am still facing
> some problem on conventions for atom names. I could not overcome
> everything still, though.
>     But in this post I would like to discuss about the best force
> fields to use in the main problems I should face: a) a protein monomer
> in water; b) a protein oligomer in water; c) a protein bound to an
> inhibitor (small organic molecule) in water; d) a protein homohexamer
> bound to DNA and co-factor in water.
>     I spent the last days gathering some information through the web
> and the gromacs manual. I could learn a lot, but I feel I still miss
> some organization of my understanding. It seems to me that specially
> for case "d" (the very one of the moment) an amber force field port
> (which best, 99SB?) to gromacs should be the preferred one, right?
> Anyway, I would like to make the question wider and ask if someone
> might point me to an article/links/whatever, for the to mature this
> overview, id est, to a compilation of the preferred (or best) force
> field for some typical biochemical MD simulations.

Answering that series of questions might take a new PhD student a month 
or more of reading, and there still won't be much in the way of 
definitive answers! You should start by finding some suitable recent 
review articles, and tracing the articles to which it refers. I don't 
have one to hand, however. Also, you need to find and read some work 
that someone has done on a similar system and see whether they came up 
with plausible simulation results. In principle, I understand CHARMM and 
AMBER forcefields are supposed to be able to deal with such systems, but 
you will need to do your own legwork there.

Mark