UBC Math Bio Seminar: Carl de Boer

The many different cell types and states result in large part from the different sets of genes they express. Gene expression level is encoded in the DNA sequence of the genome, and is interpreted by sequence-specific proteins called "transcription factors" (TFs). While characteristics of how TFs work are known, we lack a quantitative understanding of their function. Here, I describe a strategy using random DNA for building such a quantitative understanding, using yeast as a model system. I will provide a basic overview of how TFs recognize DNA, and why random DNA provides ideal gene regulatory "big data" for learning the relationship between DNA sequence and expression level. Using this strategy, we train highly complex models that learned a great deal about the biochemistry of transcriptional regulation, and gain insight into the activities of regulatory mutations.