Computational Biology at Penn

Tandy Warnow tandy at central.cis.upenn.edu
Fri Oct 14 18:28:12 EST 1994


                 PENN COMPUTATIONAL BIOLOGY SEMINAR
                   Sponsored by the Departments of
       Computer and Information Sciences, Genetics, and Biology
                  at the University of Pennsylvania


         "The Statistical Significance of Biological Sequence
                        Database Comparisons"

                       Dr. Michael S. Waterman
                      Department of Mathematics
                   University of Southern California


                     3pm Friday, October 28, 1994
                      University of Pennsylvania
                    3401 Walnut Street*, Suite 401C
                           Philadelphia, PA


One of the most important activities in computational biology is the
estimation of similarity between two sequences of DNA or protein,
presumably a reflection of their evolutionary relatedness.  A central
question in biological sequence comparison is the statistical
significance of an observed similarity, for example those that arise in
database searches.  This problem has been rigorously addressed in the
case of BLAST and related algorithms for local alignment of
subsequences allowing mismatches but not gaps due to insertions and
deletions.  However, the problem of determining the statistical
significance of optimal local alignments containing such gaps,
typically computed using well-known dynamic programming algorithms, has
so far not been solved mathematically.  A practical method will be
presented to approximate the probability that a local alignment score
is a result of chance alone.  For a set of similarity scores and gap
penalties only one simulation of random alignments needs to be
calculated to derive the key information allowing one to estimate the
significance of any alignment calculated under this setting.
Applications to database searching and the analysis of pairwise and
self-comparisons of proteins will be described.

This is the inaugural seminar of a series sponsored by the new Penn
Computational Biology Program, an interdisciplinary effort involving
the Departments of Biology, Genetics, and Computer and Information
Science.  Penn has recently been awarded a training grant by the
National Science Foundation, for PhD students and post-docs interested
in work spanning these fields.  For further information contact Dr.
Warren Ewens, at 898-7109, or by e-mail at wewens at mail.sas.upenn.edu.

*To reach the University of Pennsylvania, get onto I-76, the
Schuylkill Expressway (see below), and take exit 40 for South Street.
At the top of the ramp turn west, away from the river.  Turn right onto
33rd Street, then left onto Walnut.  The entrance to the 3401 Walnut
Street offices is on the right-hand side of Walnut, about one third
down the block west of 34th (NOT at the corner of 34th).  Take the
elevator to the 4th floor and look for the signs.  To get to I-76 from
the northeast, take the New Jersey Turnpike South to Exit 3 and the
Walt Whitman Bridge.  Approaching from the west or the north, take the
Pennsylvania Turnpike to exit 24, Valley Forge, and I-76 East.  From
the north, take I-95 South to the Center City Philadelphia exit;
follow signs to I-676 West, the Vine Street Expressway, and take the
I-76 East exit.  From the south, take I-95 North to the 291/I-76
Philadelphia exit;  after the bridge, take a left at the light
following the signs to I-76 West.



More information about the Bionews mailing list