Re: Restraint Ambiguities

From: Neema Salimi <>
Date: Thu 21 Nov 2002 21:37:13 PDT

> >
> > I was wondering if using ambiguous restraints developed for NOE
> distances
> > would be appropriate for use in enforcing hydrophobic side chain
> > contacts. For example, I define two "atoms" in the MAP file as the
> heavy
> > atoms on a certain tryptophan and a certain valine side chain and
> enforce
> > an upper limit distance constraint of 5 Angstroms. Would this be
> > effective in maintaining the contact throughout MD simulation?
> >
> This sounds correct, if I understand you. Have atom 1 be the side
> chain atoms of the trp, and atom 2 be the side chains atoms of the val.
> Then, the restraint will ensure that some atom of trp is within 5 ang. of
> some atom of val, i.e. that there is at least one such contact. If that
> sounds like what you want, give it a try.

That is what I wanted. However, the file generated by makeDIST_RST (and
a similar file from the AMBER 7 manual) brought up some questions.

First, how is the restraint "bond" actually calculated? It is based on a
flat well potential between the two centers of mass for the grouped
atoms, or is it based on individual atom positions?

Second, how does makeDIST_RST calculate the r3 (and therefore r4)
distances from the upper bound in these cases where the distance is
between groups? In my case, I defined a restraint with an upper bound of
5.5 Angstroms as a contact between any of the heavy side chain atoms
between a PHE and a TYR. makeDIST_RST created an .rst file with r3 =
10.76 Angstroms. Similarly, in a manual example (pg 136) a restraint is
defined between the amide proton and any one of the three ALA C-beta
protons. The upper bound is given as 5.5 Angstroms and r3 = 6.22
Angstroms. I guess what I am asking is does sander need r3 to equal
10.76 Angstroms to achieve the desired result, or should I edit the .rst
file to make the distance 5.5 Angstroms. Basically, this question
relates to the first question. Thanks for your help.

Neema Salimi
Graduate Student - Biophysics
University of California, San Francisco
Received on Thu Nov 21 2002 - 20:37:13 PST
Custom Search