Efficient Searches in Protein Sequence Space Through AI-Driven Iterative Learning

The protein sequence space is vast. This fact, together with the prevalence of epistasis, hampers the engineering of novel enzymes through library screening and is a major obstacle to any attempt to predict natural protein evolution. Recently, specialized methodologies have been used to determine fitness data on ~260,000 sequences for the gene of the enzyme dihydrofolate reductase and antibody affinity data for all combinations of the mutations present in the receptor-binding domain (RBD) of the Omicron strain of SARS-CoV-2 (~30,000 variants). We show that upon iterative training on a total of just a few hundred variants, various state-of-the-art AI tools (multi-layer perceptron, random forest, and XGBoost algorithms) find very high fitness variants of the enzyme and predict the antibody evasion patterns of the RBD. This work provides a basis for efficient, widely applicable, low-throughput experimental approaches to assess viral protein evolution and to engineer enzymes for biotechnological applications.