Mike's paper on sequence design of polymers by machine learning appears in Science Advances

Thursday, Oct 29, 2020

The paper "Targeted sequence design within the coarse-grained polymer genome"  is published in Science Advances. The paper demonstrates a general design approach that leverages coarse-grained modeling, supervised machine learning, and Bayesian optimization to identify polymer sequences that will yield specific properties. The strategy is demonstrated in an artificial coarse-grained chemical space where the design goal is the radius of gyration at fixed degree of polymerization. In addition to demonstrating this design paradigm, several approaches to featurizing polymer sequences are discussed. The work was picked up by a few media outlets including Materials Today and Phys.org