Re: How I can to change the force field function in NAMD for protein phi-value analysis

From: Felipe Merino (
Date: Mon Oct 07 2013 - 09:15:48 CDT


It can also be applied as a linear combination of collective variables
with PLUMED. I think you can force a residue phi value as by restraining
the fraction of native contacts you desire with a harmonic potential
(contact map is a collective variable there). If you sum each restraint
to multiplied by 1/N then you get more or less the same approach as in
the paper. You can further steer the system to 0 mean square difference
using a ratchet (a moving wall) was as in the paper as well.

The only issue is that you cannot redefine the value for the ratchet on
the basis of the trajectory of the system (eq 6 in the paper) but that
should be minor.



On 10/07/2013 03:30 PM, Giacomo Fiorin wrote:
> Hello Juan, is there a chance that the phi-value (number of contacts
> within a given cutoff) can be expressed as a /coordNum/ collective
> variable? In that case, a simple harmonic potential applied to it
> would do the trick.
> Giacomo
> On Mon, Oct 7, 2013 at 5:55 AM, Juan Jos Galano Frutos
> < <>> wrote:
> Hi dear all:
> I would like to perform MD simulations to one protein which we
> already have evaluated a series of kinetic and thermodynamic
> phi-values for a detailed conformational folding/unfolding
> analysis (transition and intermediate states included). In this
> sense, the literature suggest add a biasing energy term leading
> the system to increasingly close conformations to those suggested
> for the phi-values previously calculated.
> Such biasing pseudoenergy term is definde as (an step function):
> W(rho ,t)={[alpha * M(rho-rho0)]/2 if rho(t)> 0 , or 0 if
> rho (t)< 0} (Paci E, Vendruscolo M, Dobson CM, Karplus M (2002)
> Determination of a transition state at atomic resolution from
> protein engineering data. J Mol Biol 324:151163)
> where the parameter alpha controls the relative weight of the
> restraint term with respect to the force field, and rho0 is equal
> to the lowest value of the reaction coordinate reached by the
> system up to time t in the simulation.
> The reaction coordinate for the process, rho(t), is defined as the
> mean square difference between the phi-value(sim) and the
> experimentally determined phi-value(exp) values:
> rho(t)= 1/N * sum[phi-value(sim) - phi-value(exp)] from 1 to N,
> where N is the number of phi-value(exp) values used in the
> calculations. Meanwhile, phi-value(sim) for one conformation C in
> the simulation is defined as:
> phi-value(sim) = ni(C)/ni(native), where ni(C) is defined as
> the number of side-chain heavy atoms within a given cutoff
> distance. The ni(native) term is the same but for the native
> conformation.
> As you can see, I have clear what I would to do but I do not know
> is how in NAMD. I have some programming skills and I made some
> scripts but I'm not an expert in programming. I also researched
> and apparently this issue has explicitly not been implemented yet
> in NAMD. If someone could suggest me how to do it I would be happy.
> Thanks very much in advance.
> Juan Jos
> Ph. D Student
> Institute for Biocomputation and
> Complex Systems Physics (BIFI)
> Department of Biochemistry and
> Molecular and Cellular Biology,
> Sciences Faculty,
> University of Zaragoza
> Pedro Cerbuna # 12, 50009
> Zaragoza
> Spain
> TEL: +34 976 76 28 06 <tel:%2B34%20976%2076%2028%2006>

This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:21:45 CST