About the Seminar: 

The prediction of molecular properties plays a pivotal role in various domains, from drug discovery to materials science. With the advent of machine learning techniques, particularly in the field of cheminformatics, the prediction of properties for small organic molecules has witnessed significant advancements. This talk delves into the diverse machine-learning strategies employed for the accurate prediction of properties crucial for understanding molecular behavior. In this first part, I will highlight the development of a new radical stability metric and its utility in identifying novel radicals for aqueous redox flow batteries using graph neural networks and reinforcement learning. In the subsequent section, I will delve into the advancement of graph-based models tailored for predicting molecular properties, accompanied by an in-depth analysis of two specific applications. While the first study deals with bond dissociation energy for applications in the pharmaceutical/atmospheric, the second focuses on predicting descriptors for one of the most widely used reactions in medicinal chemistry – amide coupling reactions. In the final segment, I will showcase how computational tools can be combined to make modular workflows for FAIR (Findable, Accessible, Interoperable, and Reusable) chemistry to enable the usability and reproducibility of scientific data. By synthesizing insights from various studies, this talk aims to provide a comprehensive understanding of how machine-learning strategies are reshaping the landscape of molecular property prediction, and fostering innovation in drug design, materials science, and beyond.