Re: general question on protein-membrane simulations -complete message!

From: Luca (bellucci14_at_unisi.it)
Date: Thu Apr 19 2007 - 02:48:21 CDT

Hi Michel,
there is more possibilities for this problem
  1- Integration time step or not good r-Respa sequence for SHAKE method
       procedure ( I think that you use SHAKE rigid H-X bond)
  2- 1-4 scale factor for 1-4 interaction of same molecule
  and finally a good choice of pairlist parameter.
Did you use PME method?
Can you send these parameters?

Luca

> Sorry, I sent a truncated message. Here is the good one :)
>
> ---------------------------------------------------------------------------
>------------
>
> Hi guys,
>
> Here I am again. I'd like to share with you my previous experiences
> and I'd like to get some feedback from you. Since few months ago, I
> started working on MD simulations of protein-membrane systems. It has
> been a challenging task so far, and I still wonder how to solve an
> issue: avoid the failure of the RATTLE algorithm.
>
> I've been building my bilayers using the MEMBRANE. Then I minimize it
> and equilibrate it for about 1 ns (first applying restraints to the
> heads and finally releasing the constraints). No problem here. Then I
> combine my protein and membrane, remove overlapping lipids. I minimize
> and equilibrate (I only apply constraints to the heavy atoms of the
> protein, allowing the lipids and water to freely move; at the end,
> everything is free to move). Again, no problem here (although
> sometimes I also get it). The big problem appears when I run the
> actual production MD: after a couple of ns, I get the error related to
> the RATTLE algorithm. Usually I just restart the simulation and I got
> no problem for another 2-3 ns, let's say, and get the same error. The
> interesting thing here is that when I check the atoms that are causing
> the problem, there are not steric clashes or something anomalous. By
> restarting the simulation, I was able to get several nanoseconds, but
> at the end the RATTLE thing always appears!
>
> So my question is: Am I missing something? I think I'm properly
> equilibrating my systems, but I can't find the reasons of why I can't
> get my simulation to go straight without aborting due to RATTLE
> failure.
>
> All comments are highly appreciated.
>
> Cheers,
> Michel

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:44:35 CST