by Prof. Connor W. Coley, Massachusetts Institute of Technology, MIT (USA)
The typical molecular discovery paradigm is an iterative process of designing candidate compounds, synthesizing those compounds, and testing their performance, where each repeat of this cycle can require weeks or months, requires extensive manual effort, and relies on expert intuition. This talk will describe our efforts to accelerate the process of molecular discovery from two interrelated angles. The first involves facilitating the chemical synthesis of new structures through data-driven computer-aided synthesis planning, where we can learn patterns of chemical reactivity directly from tabulated organic reaction data. The second involves prioritizing candidate molecules to test using machine learning models for molecular property prediction. Together, these studies illustrate how data science and statistical learning offer unprecedented opportunities to systematize and streamline the process by which new functional small molecules are designed and synthesized.