Andrew Delong, Head of Computational Research
Deep Genomics
Genomics focuses on the sequences in our genomes and how they encode for function in our cells. Predicting how sequences will be interpreted by the cell is important for identifying disease-causing mutations and for designing therapies.
I will describe progress in deep learning for the prediction of protein binding and RNA splicing, which are key sequence-driven processes in the cell. I will also discuss early work on generative models that can design sequences with finely-tuned or exaggerated qualities. There are many challenges and pitfalls when working with "omics" data, so I will share some lessons learned. Time permitting, I want to give thoughts on the future of AI research in genomics and how to make the method's development more systematic and more accessible.
Bio: Andrew Delong is Head of Computational Research at Deep Genomics, a company developing AI systems for the design of genetic medicines. He was an NSERC Postdoctoral Fellow working on deep learning at the University of Toronto, where he was awarded a Heffernan Commercialization Fellowship and was co-recipient of an Invention of the Year Award for his work in machine learning for genomics. His other research interests include computer vision and optimization algorithms.
He received the B.Math degree in computer science from the University of Waterloo. He received the M.Sc. and Ph.D. degrees in computer science from Western University, receiving a Dissertation Award from the Canadian Image Processing and Pattern Recognition Society for his work in computer vision. He previously worked in the computer graphics industry on what is now Autodesk Maya.