Benchmarking deep generative models for diverse antibody sequence design

TL;DR

This paper benchmarks three deep generative models for antibody sequence design, highlighting Fold2Seq's superior ability to generate diverse sequences while preserving structural integrity.

Contribution

It introduces a comprehensive comparison of recent deep generative frameworks for protein design, emphasizing the effectiveness of Fold2Seq in antibody sequence diversity.

Findings

Fold2Seq outperforms other models in sequence diversity.

Fold2Seq maintains structural consistency in designed sequences.

Benchmark results guide future model development in protein design.

Abstract

Computational protein design, i.e. inferring novel and diverse protein sequences consistent with a given structure, remains a major unsolved challenge. Recently, deep generative models that learn from sequences alone or from sequences and structures jointly have shown impressive performance on this task. However, those models appear limited in terms of modeling structural constraints, capturing enough sequence diversity, or both. Here we consider three recently proposed deep generative frameworks for protein design: (AR) the sequence-based autoregressive generative model, (GVP) the precise structure-based graph neural network, and Fold2Seq that leverages a fuzzy and scale-free representation of a three-dimensional fold, while enforcing structure-to-sequence (and vice versa) consistency. We benchmark these models on the task of computational design of antibody sequences, which demand designing sequences with high diversity for functional implication. The Fold2Seq framework outperforms the two other baselines in terms of diversity of the designed sequences, while maintaining the typical fold.