mm-pbsa problem & possible solution

From: Thomas Steinbrecher <thomas.steinbrecher_at_physchem.uni-freiburg.de>
Date: Thu 30 Jan 2003 16:51:57 +0100

Dear AMBER users,

We have used the MM-PBSA script to process AMBER7
trajectories and found that the script fails when the
coordinates contain lines like:

ATOM 4136 CB GLN 253 -2.782-100.885 86.358

where the x and y coordinate columns are separated by a
minus sign and no whitespaces.

After some error hunting we presume the problem lies in
line 1316 of mm_pbsa.pl:

      $line =~ /(ATOM +|HETATM +)(\d+ +)(\S+)( +)(\S+)(
 +\d+)( +-?\d+\.\d+ +-?\d+\.\d+ +-?\d+\.\d+)/;

because it expects the coordinate columns to be seperated
by at least one whitespace and zero or more minus signs.

We think that changing the line to:

      $line =~ /(ATOM +|HETATM +)(\d+ +)(\S+)( +)(\S+)(
 +\d+)( +-?\d+\.\d+[ \-]+\d+\.\d+[ \-]+\d+\.\d+)/;

solves the problem.

Has anybody experienced the same problem with MM_PBSA runs?
I am no perl expert, so please correct me if the changes
made are not reasonable.

Kind Regards,

Thomas Steinbrecher
Received on Thu Jan 30 2003 - 07:51:57 PST
Custom Search