Using Artificial Intelligence to Model and Understand Genetic Disease

About this Class

Predicting how changes in the genome manifest as phenotypic differences is an extremely challenging problem that requires a deep understanding of multiscale biological mechanisms. And while we know a great deal about how information stored in a sequence of nucleotides translates into the complexities of life, our understanding of how subtle changes on the molecular scale can lead to drastic changes in phenotype is incomplete. In the age of genomic sequencing and the wealth of information on variation in the human genome, predicting the degree a variant of unknown significance will contribute to the pathogenicity of a disease is a challenge that can only be addressed by a computational approach. And not just because the prevalence of genomic variation makes experimental characterization intractable, advances in Artificial Intelligence (AI) provide a means to learn the multiscale complexity and emergent properties that drive genetic disease. Our preliminary studies have shown AI trained on simulations of variant protein dynamics can segregate between related but distinct disease mechanisms, and is even predictive of disease severity. As our knowledge of variant-disease associations continues to grow, AI models that connect variation in DNA to disease phenotypes will become an integral part of how we understand, assess, and treat genetic disease. Speaker: Matthew McCoy, Ph.D., Assistant Professor, Department of Oncology, Georgetown University Medical Center Biography: - PhD, Bioinformatics and Computational Biology, George Mason University - Assistant Professor, Department of Oncology, Georgetown University Medical Center - Contributes to the research and education mission of Georgetown University's Innovation Center for Biomedical Informatics - Research interests: Using the information gleaned through various high throughput technologies to parameterize physiologically realistic, multi-scale models of biological systems, with the ultimate goal of informing therapeutic decision making though personalized models of genetic disease. - Accolades: Received the Marco Ramoni Distinguished Paper Award for work he presented at the AMIA 2018 Informatics Summit.