Mathematics Colloquia and Seminars

Return to Colloquia & Seminar listing

Description

ABSTRACT: The study of molecular biology on a whole genome scale has only recently become possible as a result of the many ongoing genome sequencing projects. In order to understand the relationship between sequences and their biological function it is helpful to accurately annotate the sequences, but this problem has proved to be very difficult. Another approach for extracting information from the sequences is to compare and contrast them. In this way it is possible to infer information about a newly sequenced organism by extrapolating from what is known about another (this is similar to the way the Rosetta stone was used to decipher Egyptian hieroglyphics using ancient Greek). We will review the mathematical tools used for similarity searching (alignment), and for annotation (gene recognition). We will then show that hidden Markov models provide a probabilistically sound theory for unifying and generalizing classical results about the two problems.