EvoFlows: Evolutionary Edit-Based Flow-Matching for Protein Engineering
This addresses protein engineering for researchers by offering a novel method to generate mutants, though it is incremental as it builds on existing flow-matching and protein modeling techniques.
The paper tackles protein engineering by introducing EvoFlows, a variable-length sequence-to-sequence model that performs controllable insertions, deletions, and substitutions on template proteins, achieving quality comparable to leading masked language models and improved generation of natural-like mutants.
We introduce EvoFlows, a variable-length sequence-to-sequence protein modeling approach uniquely suited to protein engineering. Unlike autoregressive and masked language models, EvoFlows perform a limited, controllable number of insertions, deletions, and substitutions on a template protein sequence. In other words, EvoFlows predict not only _which_ mutation to perform, but also _where_ it should occur. Our approach leverages edit flows to learn mutational trajectories between evolutionarily-related protein sequences, simultaneously modeling distributions of related natural proteins and the mutational paths connecting them. Through extensive _in silico_ evaluation on diverse protein communities from UNIREF and OAS, we demonstrate that EvoFlows capture protein sequence distributions with a quality comparable to leading masked language models commonly used in protein engineering, while showing improved ability to generate non-trivial yet natural-like mutants from a given template protein.