Date: Tue Mar 27 2007 - 19:59:03 CDT
With a lot of help from Gianluca Interlandi, we have identified the
cause of this problem. I was adding the caps to the peptides as patches
(ACE and CT2) rather than as "first" and "last". This resulted in no
angle restraints at all for the caps in the psf file made with psfgen -
and then not surprisingly the geometry of the caps became very strange
during subsequent minimization.
From: owner-namd-l_at_ks.uiuc.edu [mailto:owner-namd-l_at_ks.uiuc.edu] On
Behalf Of Nairn, Kate (CMMT, Clayton)
Sent: Tuesday, 6 March 2007 2:18 PM
Subject: namd-l: Difficulties is minimizing capped peptides
I am attempting to simulate 10-residue peptides in explicit water with
salt. I can perform minimization, equilibration and dynamics runs on
peptides without caps. But when I try to carry out similar processes on
peptides with capped termini, things get strange.
The caps are applied using the following patch residues in PSFGEN:
N-terminal ACE and C-terminal CT2
The initial pdb/psf files look fine. The problems arise early on in the
minimization process, where both the methyl group in the ACE cap and the
NH2 group of the CT2 cap turn "inside out", with the hydrogen atoms all
pointing back towards the main chain. Eventually the total energy does
go to a minimum, the gradient tolerance decreases to around 4 after
1000 steps, but these "inside-out" groups remain.
Any attempt to subsequently equilibrate these structures results in
"constraint failure in rattle algorithm" early in the equilibration
(either in the 0-25K step or 25-50K step).
Other than manually pulling the atoms back closer to where they should
be, or not minimizing the structure in the first place, are there any
ways of making ACE and CT2 caps behave better?
Thanks in advance
Research Scientist - Biopolymer Characterisation
CSIRO Manufacturing and Materials Technology
Gate 5, Normanby Rd, Clayton, 3168
phone :+61 3 9545 2667
fax: +61 3 9544 1128
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:44:30 CST