Geoffrey Siwo

My research is driven by a strong belief that abstracting complex scientific problems into computational problems has the power to accelerate the speed of scientific discovery and change who, here and how scientific discoveries are made and translated into practical solutions. I apply this principle to integrate computational techniques such as machine learning and artificial intelligence with fundamental biological knowledge and data at different scales – from the molecular to the phenotypic- to advance basic understanding of human biology, the design of therapeutics and public health interventions. A specific focus area of my current work includes developing computational frameworks that leverage large language models (LLMs) to encode and reason over knowledge at different scales and from different domains. Applied projects I am conducting in this area include i) developing computational frameworks to bridge the gap between fundamental biological knowledge, high-throughput biological datasets and electronic health records to advance biological discoveries especially for rare diseases; ii) automated design of RNA therapeutics and, iii) enhancing the utilization of scientific evidence by non-scientists. I am a strong proponent of leveraging frameworks for scientific collaborations such as open innovation challenges and hackathons that break barriers across disciplines, geographies and level of expertise to explore new ideas in unconventional ways. I serve as organizer and in committees for such challenges including the NeuRIPS Competitions Track.