A Variational Perspective on Generative Protein Fitness Optimization

TL;DR

This paper introduces VLGPO, a variational approach that embeds protein sequences in a continuous space, enabling efficient optimization of protein fitness with state-of-the-art results and flexible customization.

Contribution

It presents a novel variational framework for protein fitness optimization that combines a learned prior and fitness predictor in a continuous latent space.

Findings

Achieved state-of-the-art results on multiple protein benchmarks.

Demonstrated flexible and customizable protein design framework.

Abstract

The goal of protein fitness optimization is to discover new protein variants with enhanced fitness for a given use. The vast search space and the sparsely populated fitness landscape, along with the discrete nature of protein sequences, pose significant challenges when trying to determine the gradient towards configurations with higher fitness. We introduce Variational Latent Generative Protein Optimization (VLGPO), a variational perspective on fitness optimization. Our method embeds protein sequences in a continuous latent space to enable efficient sampling from the fitness distribution and combines a (learned) flow matching prior over sequence mutations with a fitness predictor to guide optimization towards sequences with high fitness. VLGPO achieves state-of-the-art results on two different protein benchmarks of varying complexity. Moreover, the variational design with explicit prior and likelihood functions offers a flexible plug-and-play framework that can be easily customized to suit various protein design tasks.