Ai+ Training

High uncertainty is inherent to disease treatment and diagnosis. Despite significant advances in basic biology and clinical science, we still cannot predict which individuals are likely to get sick, whether their disease will be treatable and what side effects they will experience during and post their treatment. AI models have potential to transform this centuries-old status quo due to their uncanny capacity to predict future outcomes. In this talk, I will explain how these models make their predictions and illustrate it with specific examples in the area of risk assessment and cancer diagnosis. In the second part of my talk, I will focus on AI based methods for drug discovery, including generative models for drug design and mechanism of action elucidation.