Learning Expressive Models of Gene Regulation

A central challenge in computational biology is to uncover the mechanisms and cellular circuits that govern how the expression of various genes is controlled in response to a cell's environment. This challenge presents many interesting opportunities for machine-learning methods, especially those that employ expressive representations. In this talk, I will discuss recent research in using machine-learning approaches to (i) recognize regulatory elements in genomic sequences, (ii) uncover networks of interactions among genes, and (iii) characterize the cellular responses induced by various stimuli. I will highlight tasks that call for models that use expressive representations, and discuss lessons learned about what types of representational attributes are important for these tasks.