Machine learning-assisted directed protein evolution with combinatorial libraries

Significance Proteins often function poorly when used outside their natural contexts; directed evolution can be used to engineer them to be more efficient in new roles. We propose that the expense of experimentally testing a large number of protein variants can be decreased and the outcome can be improved by incorporating machine learning with directed evolution. Simulations on an empirical fitness landscape demonstrate that the expected performance improvement is greater with this approach. Machine learning-assisted directed evolution from a single parent produced enzyme variants that selectively synthesize the enantiomeric products of a new-to-nature chemical transformation. By exploring multiple mutations simultaneously, machine learning efficiently navigates large regions of sequence space to identify improved proteins and also produces diverse solutions to engineering problems. To reduce experimental effort associated with directed protein evolution and to explore the sequence space encoded by mutating multiple positions simultaneously, we incorporate machine learning into the directed evolution workflow. Combinatorial sequence space can be quite expensive to sample experimentally, but machine-learning models trained on tested variants provide a fast method for testing sequence space computationally. We validated this approach on a large published empirical fitness landscape for human GB1 binding protein, demonstrating that machine learning-guided directed evolution finds variants with higher fitness than those found by other directed evolution approaches. We then provide an example application in evolving an enzyme to produce each of the two possible product enantiomers (i.e., stereodivergence) of a new-to-nature carbene Si–H insertion reaction. The approach predicted libraries enriched in functional enzymes and fixed seven mutations in two rounds of evolution to identify variants for selective catalysis with 93% and 79% ee (enantiomeric excess). By greatly increasing throughput with in silico modeling, machine learning enhances the quality and diversity of sequence solutions for a protein engineering problem.