[gmx-users] Umbrella sampling - chain A does not get pulled
jalemkul at vt.edu
Tue Dec 11 14:38:50 CET 2012
On 12/11/12 2:40 AM, Davide Mercadante wrote:
> Dear Justin,
> I have been practicing umbrella sampling simulations following your tutorial
> step by step. I have just finished to perform the pull simulations to
> identify the configurations to use in the umbrella runs. I have used the
> distances.pl script to run iteratively g_dist and the resulting file
> (summary_distances.dat) is showing a distance between 2.62nm (for conf0.gro)
> and 5.3nm (for conf500.gro). I don't seem to have the values reported in the
> tutorial at all (in the tutorial is reported that the COM distance would be
> something of an order of magnitude lowerŠ) and I am not sure if something is
> gone wrong.
The distances shown in the tutorial are just an illustrative example and don't
necessarily correspond to anything you might obtain. A distance of 2.62 nm for
the initial configuration does not seem right to me though; it should be on the
order of 0.5 nm, coincident with interpeptide spacing in the beta-sheet.
> I have also tried to visualize the trajectories but I am not able to see
> chain_A moving away from the rest of the fibril. I have set, at the end of
> the topol_Protein_chain_B.itp file, the lines suggested to restrain chain B.
> #ifdef POSRES_B
> #include "posre_Protein_chain_B.itp"
> I have used all the input files given in the tutorial and the commands/files
> extensions suggested.
> Can you please help me to understand if I am doing something wrong?
What you're reporting is not consistent - your distances increase over time, but
then you say peptide A does not move away from the others? That's not possible.
If peptide A doesn't move, then the COM distance should remain relatively
constant. Are you loading the right trajectory (i.e., the actual SMD, not a
previous equilibration or something else)?
Justin A. Lemkul, Ph.D.
Department of Biochemistry
jalemkul[at]vt.edu | (540) 231-9080
More information about the gromacs.org_gmx-users