Thursday, March 10, 2005
4:15 pm
B17 Upson Hall

Computer Science
Spring 2005

Michael Brudno
UC Berkeley

Alignment of Whole Genomes: Techniques and Algorithms

One of the most fruitful ways to identify the functional areas of a genome is to compare the sequences of several related organisms. Because the functional areas are under constraint, they change more slowly over evolutionary time, and hence are usually conserved in related genomes. In order to find these functional sequences biologists need alignment methods that are efficient enough to handle long sequences, and accurate enough to correctly align the conserved biological features between distant species.

In particular, the availability of several mammalian genomes necessitates the development of tools for multiple alignment of large genomes. In this talk I will describe a method for multiple alignment of several long sequences, an approach to aligning two genomic sequences in the presence of rearrangements, and also present different methods for multiple alignment of whole genomes. I will also demonstrate how using these alignment algorithms biologists have been able to better understand the evolutionary forces that affect a genome and also to characterize many functional elements in the mammalian genome.