AMBER Archive (2003)

Subject: How do I perform an Nmode calculation on a big protein-ligand complex?

From: Thomas Steinbrecher (thomas.steinbrecher_at_physchem.uni-freiburg.de)
Date: Mon Jun 30 2003 - 04:31:38 CDT

Dear AMBER users,

I am trying to estimate the change of entropy upon binding
of a ligand to a receptor as part of an MM-PBSA calculation
using AMBER7 on Linux.

What I did so far is to take snapshots of my complex from
an MD an minimize them using sander.

Since my complex is to big to be treated by nmode as a
whole for normal mode analysis, I plan to use the approach
outlined in the paper by Kuhn and Kollman (J. Med. Chem.
2000, 43, 3786-3791).

They were using size-reduced snapshots (cutouts of 8A
around the ligand) for their nmode calculations and I
wonder how to create the parmtop and coordinate files for
such a size-reduced snapshot.

I tried Carnal, which can give me a list of residues that
make up an 8A sphere around my ligand, but no parameter or
coordinate files.
If I generate a pdb-file of a size-reduced snapshot, I
cannot read it into leap without leap adding terminal
oxygens.
I also wonder how valid a normal mode calculation is in
which only unconnected peptide fragments of my receptor
surround the ligand.

Can anybody give me some hints on how to do such a
calculation with the programs in the amber suite?

Many thanks in advance,

Thomas Steinbrecher