From: Sébastien Légaré (Sebastien.Legare_at_rsvs.ulaval.ca)
Date: Tue Oct 28 2008 - 11:50:02 CDT
I am trying to use ABF to compute the free energy involved in a simple
conformational change of a small peptide. The structure of the peptide
presents an N terminal helix, followed by an hinge at one precise residue and
a C terminal helix. The conformational change highly correlates with the
rotation of the hinge residue backbone phi dihedral angle. My goal is to
compute the free energy of rotation of this dihedral angle while all the
other backbone dihedral angles are kept at the caracteristic value of alpha
At first I thought I would use dihedral.tcl on the hinge phi angle and apply
restraints on all other backbone dihedrals. This implies that the atoms
involved in the RC will also be part of some restraints, which should not be
done according to the user's guide.
My first question is : Could the limitation of having no restraint on RC atoms
be overcome in future versions of ABF or does it come from the physics?
Instead, I tried by modifying the dihedral.tcl file to compute the total
torque applied on both helixes (contributions from each atoms of an helix)
around the axis defined by the phi dihedral central atoms. I then apply the
inverse torque back on the system in the (I think) correct way to keep the
helical structures, see attached dihedral-helix.tcl file. Up to now, it could
not help my system to go through its energy barrier. The system either stays
near the free energy minimum or gets pushed toward the nearest end of the RC
(where the free energy is high) and stays there. Already seen this problem?
If anyone have some time to spend, I would like to know if the general idea
behind my script is correct. I put a lot of comments.
Thank you very much
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:50:01 CST