14. Georg Seelig: Machine-learning guided sequence design for mRNA and gene therapy

Abstract. Predicting the impact of cis-regulatory sequence on gene expression is a foundational challenge for biology. In this talk, I will present our work on building models that predict molecular phenotypes (e.g. transcription, translation, stability) from gene sequence. Computational models are often trained on very large reporter gene libraries with millions of members and containing random sequences. Crucially, such sequence-function models can generalize from training data to unseen sequences by learning the regulatory rules underlying the observed molecular phenotype. The massive size of the (synthetic) training data allows us to improve upon models trained exclusively on genomic data, even on the task of predicting the impact of human genetic variation on gene function. When combined with sequence design algorithms, models can be used to generate functional cis-regulatory sequences. We apply this approach to design enhancers that result in cell type specific gene expression, and mRNA UTR sequences that result in high levels of translation or stability and that can find applications in mRNA and gene therapy.
Bio. Georg Seelig is a professor in the Department of Electrical & Computer Engineering and the Paul G. Allen School of Computer Science & Engineering at the University of Washington. He is an adjunct professor in Bioengineering. The Seelig group is interested in understanding how biological organisms process information using complex biochemical networks and how such networks can be engineered to program cellular behavior. Seelig holds a PhD in physics from the University of Geneva in Switzerland and did postdoctoral work in synthetic biology and DNA nanotechnology at Caltech. He received a Burroughs Wellcome Foundation Career Award at the Scientific Interface, an NSF Career Award, a Sloan Research Fellowship, a DARPA Young Faculty Award, an ONR Young Investigator Award and a Rozenberg Tulip Award in DNA computing among others.