[gmx-users] Re:Peptide and bilayer sims

[Senthil Kandasamy]

Though Graham Smith is hesitant to suggest his hole program, I highly
recommend it to insert short peptides in lipid bilayers. I have been
using the hole program to insert short antimicrobial peptides in lipid
monolayers and bilayers in various regions of the head and tail groups.
It has been working really well for me. I have had to use really large
values of the hole-making force constant, but have been following it up
with atleast 1-2 ns of position-restrained equilibration to let the
highly distorted lipids re-equilibrate before I start the production
run. 

But as Graham mentioned, the initial conformation of the peptide that
one chooses is very important. I have been using structures obtained
from NMR data in hydrophobic environments,  which would presumably be
close to the membrane associated structure. However, from simulations in
the 10-20 ns timescale, I do not observe drastic rearrangement of the
peptide.The propensity of the peptide to get stuck in local minima is
very high and to really observe peptide insertion, one might have to do
very long simulations or use something like replica exchange md. 
Nevertheless,I think that this is a difficult and important problem and
I wish more people would work on it.