Transformer-based protein generation with regularized latent space optimization

The development of powerful natural language models has improved the ability to learn meaningful representations of protein sequences. In addition, advances in high-throughput mutagenesis, directed evolution and next-generation sequencing have allowed for the accumulation of large amounts of labelled fitness data. Leveraging these two trends, we introduce Regularized Latent Space Optimization (ReLSO), a deep transformer-based autoencoder, which features a highly structured latent space that is trained to jointly generate sequences as well as predict fitness. Through regularized prediction heads, ReLSO introduces a powerful protein sequence encoder and a novel approach for efficient fitness landscape traversal. Using ReLSO, we explicitly model the sequence–function landscape of large labelled datasets and generate new molecules by optimizing within the latent space using gradient-based methods. We evaluate this approach on several publicly available protein datasets, including variant sets of anti-ranibizumab and green fluorescent protein. We observe a greater sequence optimization efficiency (increase in fitness per optimization step) using ReLSO compared with other approaches, where ReLSO more robustly generates high-fitness sequences. Furthermore, the attention-based relationships learned by the jointly trained ReLSO models provide a potential avenue towards sequence-level fitness attribution information. The space of possible proteins is vast, and optimizing proteins for specific target properties computationally is an ongoing challenge, even with large amounts of data. Castro and colleagues combine a transformer-based model with regularized prediction heads to form a smooth and pseudoconvex latent space that allows for easier navigation and more efficient optimization of proteins.