Re: Changing loop to helix: restraining about 20 dihedral angles

From: Peter Freddolino (petefred_at_ks.uiuc.edu)
Date: Thu Jun 21 2007 - 09:06:00 CDT

If all you're attempting to do is get the helix formed the easiest and
fastest way to do it currently is probably to invoke the abf module and
apply dihedral constraints through it; these would be harmonic
constraints. If you're feeling more adventurous, the CVS version of namd
includes a new feature (extraBonds) that allows you to specify such
constraints directly and have them calculated along with other
interactions of the applied type; going that route would require getting
a cvs account and messing with some new features. Option 3 is also a
good one, although you don't seem to like it...

As far as speed is concerned, options 1 and 2 will have roughly the same
speed since they're doing the force calculations in tcl and not in the
(parallelized) namd bonded force evaluation routines. Option 3 and
extraBonds should be faster than 1 and 2, and will be equivalent to each
other.

Peter

maria goranovic wrote:
> Er.. anyone ?? Help !!!
>
>
>
> On 6/19/07, *maria goranovic* <mariagoranovic_at_gmail.com
> <mailto:mariagoranovic_at_gmail.com>> wrote:
>
> I want to restrain about 20 random coil contiguous residues to a
> helical conformation, while leaving the rest of my protein intact.
> According to the documentation, there are at least three ways to
> do this:
>
> 1. Using a tcl forces script, like the one recommended in the
> manual.
> http://www.ks.uiuc.edu/Research/namd/2.6b1/ug/node32.html#SECTION00096700000000000000.
> However, the script may get slightly tedious for 20 operations.
> 2. Using the restraining options in the free energy methods. But
> here, the free energy routines will need to be invoked, which
> might slow the overall calculation, because an unnecessary
> derivative will be calculated each step.
> 3. Modify the psf to define new residue types and new dihedrals (I
> want to avoid this).
>
> My questions:
>
> a. Which is the recommended method in terms of saving of
> computational time ?
> b. In the free energy method, is the dihedral restrained
> harmonically throughout the simulation like it would be in the TCL
> forces script ?
> c. In terms of added potentials and forces, is the effect of using
> a tcl forces script or a restraint in a free energy calculation
> similar to that, say of using standard harmonic constraints in NAMD ?
>
> My goal is to investigate the effect of the coil-helix
> transformation on the loop region next to the protein region in
> question.
>
> Any suggestions please ? All partial answers are also welcome :)
>
> -Maria
>
> --
> Maria G.
> Technical University of Denmark
>
>
>
>
> --
> Maria G.
> Technical University of Denmark
> Copenhagen

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:44:52 CST