Reinforcement-guided generative protein language models enable de novo design of highly diverse AAV capsids

TL;DR

This paper introduces a reinforcement learning-guided generative framework using protein language models to design highly diverse and viable AAV capsid sequences, expanding the exploration of protein sequence space for gene therapy applications.

Contribution

The study develops a novel reinforcement learning-based method combined with protein language models for de novo AAV capsid design, enabling exploration beyond the training data while maintaining viability.

Findings

Reinforcement learning guides generation towards more novel sequences.

Fine-tuning alone biases sequences towards known data.

The framework effectively balances viability and novelty in candidate selection.

Abstract

Adeno-associated viral (AAV) vectors are widely used delivery platforms in gene therapy, and the design of improved capsids is key to expanding their therapeutic potential. A central challenge in AAV bioengineering, as in protein design more broadly, is the vast sequence design space relative to the scale of feasible experimental screening. Machine-guided generative approaches provide a powerful means of navigating this landscape and proposing novel protein sequences that satisfy functional constraints. Here, we develop a generative design framework based on protein language models and reinforcement learning to generate highly novel yet functionally plausible AAV capsids. A pretrained model was fine-tuned on experimentally validated capsid sequences to learn patterns associated with viability. Reinforcement learning was then used to guide sequence generation, with a reward function that jointly promoted predicted viability and sequence novelty, thereby enabling exploration beyond regions represented in the training data. Comparative analyses showed that fine-tuning alone produces sequences with high predicted viability but remains biased toward the training distribution, whereas reinforcement learining-guided generation reaches more distant regions of sequence space while maintaining high predicted viability. Finally, we propose a candidate selection strategy that integrates predicted viability, sequence novelty, and biophysical properties to prioritize variants for downstream evaluation. This work establishes a framework for the generative exploration of protein sequence space and advances the application of generative protein language models to AAV bioengineering.