Brendan J. Frey
Electrical and Computer Engineering
University of Toronto
Construction of the discrete genome sequence was the first step in developing a comprehensive understanding of how cellular processes are controlled by bio-molecules and their interactions. That step is now mostly complete and the next step is to determine how DNA subsequences encode instructions for producing RNA transcripts and how continuous abundances of transcripts in cells combine to control activities. This is a much more challenging task than genome assembly, because the encoding of genetic instructions turns out to be far richer than was previously thought, and the detection and analysis of continuous cellular signals is more difficult than discrete symbol detection. Only preliminary progress has been made in assembling and analyzing the "transcriptome" and the first genome-wide data sets enabling the study of transcripts and their interactions have only recently been published. In this talk, I'll describe several open research problems in this area and discuss how they can be approached using representations and algorithms familiar to researchers in the information theory community.