PIMA bug fix

Randall Smith rsmith at kite.bcm.tmc.edu
Wed Feb 19 10:20:27 EST 1992

BUG FIX FOR PIMA-1.2     (Pattern-Induced Multi-sequence Alignment program)
        AND PLSEARCH-4.1 (MBCRR's Pattern Library and Search tool)

During the recent upgrade of pima-1.0 to pima-1.2 and plsearch-4.0 to
plsearch-4.1 a bug was introduced into the alignment algorithm that
affects the alignment of gaps across from long runs of gap characters
(a rare enough occurrence such that it was not detected during program
testing, but a serious enough bug that users should pick-up the fix).

Corrected versions of PIMA and PLSEARCH (pima-1.22 and plsearch-4.2)
are now available via anonymous ftp from mbcrr.harvard.edu
(both packages are in the 'MBCRR-Package' sub-directory as compressed
tar files; be sure to use 'binary' mode when transferring).

| Randall F. Smith                           | Phone: 713-798-4735          |
| Keck Center for Computational Mol. Biology | FAX: 713-790-1275            |
| and Dept. of Cell Biology                  |                              |
| Baylor College of Medicine                 |                              |
| One Baylor Plaza                           | Internet: rsmith at bcm.tmc.edu |
| Houston, TX  77030  USA                    |                              |

The original PIMA-1.2 announcement follows:

               Announcing Rel. 1.2 of PIMA

     Pattern-Induced Multi-sequence Alignment Program

PIMA is a multiple sequence alignment program that uses a dynamic
programming-based   pattern  construction  method to align a  set
of  homologous protein  sequences  (Smith,   RF  and  Smith,  TF,
1992, Protein Engineering vol 5, num 1 (should be in US libraries
this week; see also Smith and Smith, 1990, PNAS 87:118-122).

The multiple alignment algorithm:

        (1)  is  based  on   a   local  (Smith-Waterman)   rather
than   global  (Needleman-Wunch-Sellars) dynamic programming algo-

        (2) aligns sequences based on the  pattern  of  conserved
sequence  elements/domains common to the homologous sequences be-
ing aligned;

        (3)  represents gaps  within  a  multiple alignment in  a
simple and consistent manner;

        (4) can employ secondary structure-dependent gap   penal-
ties  for  use  in  structural  modeling of new protein sequences
when the 3D-structure of one or more members of the sequence fam-
ily is known.

The PIMA package is available free of charge to non-profit organ-
izations; a distribution fee and non-resale agreement is required
for commercial use.

Note: The PIMA package is currently a Unix-only implementation.

Copies  can  be  obtained  via   INTERNET   anonymous   ftp   to:
mbcrr.harvard.edu  =;  the  package is in a single
compressed tar file called pima-1.2.tar.Z  in  the  MBCRR-Package
sub-directory, and

Note: Temple and I have both moved recently:

Randall F. Smith
Human Genome Center and Dept. of Cell Biology
Baylor College of Medicine, Houston TX  77096
rsmith at bcm.tmc.edu

Temple F. Smith
Molecular Bio-Enginnering Research Center
Boston Univ.,  36 Cummington St, Boston, MA 02115
tsmith at darwin.bu.edu

More information about the Bio-soft mailing list