Transferable coarse-grained models accelerate chemical-space exploration

T. Bereau

Advanced statistical methods are rapidly impacting many scientific fields, offering new perspectives on long-standing problems. In materials science, data-driven methods are already bearing fruit in various disciplines, such as hard condensed matter or inorganic chemistry, while comparatively little has happened in soft matter. I will describe how we use multiscale simulations to leverage data-driven methods in soft matter. We aim at establishing structure-property relationships for complex thermodynamic processes across the chemical space of small molecules. Akin to screening experiments, we devise a high-throughput coarse-grained simulation framework. Coarse-graining is an appealing screening strategy for two main reasons: it significantly reduces the size of chemical space and it can suggest a low-dimensional representation of the structure-property relationship. I will describe how we accelerate molecular discovery by blending representation learning, free-energy calculations, and a Bayesian-optimization framework. We used this framework in the context of a complex biomolecular problem that led to the discovery of in vivo active compounds. Finally, I will discuss how exploiting the hierarchical nature of coarse-grained models can further accelerate the exploration of chemical space.