- This event has passed.
Department of Bioengineering Seminar Series-Probing the sequence-to-function relationship of transcription factor activation domains
March 9 @ 12:00 pm - 1:00 pm
To activate transcription, transcription factors require a DNA binding domain (DBD) and one or more activation domains. DBDs are well understood compared to activation domains: DBDs are conserved, structured and can be predicted directly from amino acid sequence. Activation domains are poorly conserved, intrinsically disordered and, consequently, cannot be predicted from amino acid sequence. To work towards building computational models for predicting activation domains from amino acid sequence, we have developed high throughput methods for studying thousands of activation domain variants in parallel in yeast and human cell culture. We investigated the amino acid features that contribute to activation domain activity. Acidic activation domains were first described over thirty years ago, but it always been unclear why they are acidic. Based on our work in yeast, we proposed a unified model where acidity and intrinsic disorder keep key aromatic residues exposed to solvent, where they can interact with coactivators. We have recently shown that this model explains the effects of mutations in multiple human activation domains. These large-scale data, coupled with machine learning, will lay a foundation for computational models that predict activation domains from amino acid sequence.
Max began his interdisciplinary training as a member of the first cohort of the Integrated Science Curriculum at Princeton University. As PhD student in Systems Biology at Harvard University, he worked with Angela DePace to build and experimentally test computational models for how fruit fly embryos build segments. As a postdoc at Washington University working with Barak Cohen, he is studying the sequence features of transcription factor activation domains.