Extent of Sampling possible using classical all-atom MD Simulations

From: Aditya Ranganathan (aditya.sia_at_gmail.com)
Date: Tue Oct 01 2013 - 08:22:42 CDT

Dear All,

Here is a question that has been troubling a few in our lab. We are trying
to study the conformational switching of a monomeric highly folded protein
(approx 500 residues long) and were wondering if it is actually possible to
sample the multiple states accessed by the protein with classical all atom
molecular dynamics provided we can run long enough (lets assume we have
enough computing power to run long enough simulations). Are all atom MD
simulations in principle powerful enough to answer questions of this kind
or is it possible for the proteins to remain kinetically trapped forever in
some state (say 'state A'), unable to switch to state to state B even if I
run a simulation for an infinitely long timescale. If so, why have replica
exchange molecular dynamics simulations become more common in answering
such conformational switching questions in protein structural biology? Is
it solely because of computing requirements of due to possible kinetic
trapping for the proteins in some state.

Are there any classical literature pertaining to the same question?


Srivastav Ranganathan
Reserach Scholar
Department of Biosciences and Bioengineering,
IIT Bombay, Mumbai

This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:21:44 CST