Abstract:Understanding the relationships between protein sequence, structure and function is a long-standing biological challenge with manifold implications from drug design to our understanding of evolution. Recently, protein language models have emerged as the preferred method for this challenge, thanks to their ability to harness large sequence databases. Yet, their reliance on expansive sequence data and parameter sets limits their flexibility and practicality in real-world scenarios. Concurrently, the recent surge in computationally predicted protein structures unlocks new opportunities in protein representation learning. While promising, the computational burden carried by such complex data still hinders widely-adopted practical applications. To address these limitations, we introduce a novel framework that enhances protein language models by integrating protein structural data. Drawing from recent advances in graph transformers, our approach refines the self-attention mechanisms of pretrained language transformers by integrating structural information with structure extractor modules. This refined model, termed Protein Structure Transformer (PST), is further pretrained on a small protein structure database, using the same masked language modeling objective as traditional protein language models. Empirical evaluations of PST demonstrate its superior parameter efficiency relative to protein language models, despite being pretrained on a dataset comprising only 542K structures. Notably, PST consistently outperforms the state-of-the-art foundation model for protein sequences, ESM-2, setting a new benchmark in protein function prediction. Our findings underscore the potential of integrating structural information into protein language models, paving the way for more effective and efficient protein modeling Code and pretrained models are available at

Distilling structural representations into protein sequence models

ESM All-Atom: Multi-scale Protein Language Model for Unified Molecular Modeling

Efficient Inference, Training, and Fine-tuning of Protein Language Models

Structure Language Models for Protein Conformation Generation

Beyond ESM2: Graph-Enhanced Protein Sequence Modeling with Efficient Clustering

Endowing Protein Language Models with Structural Knowledge

A Systematic Study of Joint Representation Learning on Protein Sequences and Structures

Multi-Scale Protein Language Model for Unified Molecular Modeling

Structure-Informed Protein Language Model

SFM-Protein: Integrative Co-evolutionary Pre-training for Advanced Protein Sequence Representation

SaProt: Protein Language Modeling with Structure-aware Vocabulary

Protein complex structure modeling by cross-modal alignment between cryo-EM maps and protein sequences

Protein language models learn evolutionary statistics of interacting sequence motifs

More Structures, Less Accuracy: ESM3's Binding Prediction Paradox

Learning protein sequence embeddings using information from structure

Structure-Enhanced Protein Instruction Tuning: Towards General-Purpose Protein Understanding

CPE-Pro: A Structure-Sensitive Deep Learning Model for Protein Representation and Origin Evaluation

Structure-informed Language Models Are Protein Designers

Exploring evolution-aware & -free protein language models as protein function predictors

Evolutionary-scale prediction of atomic-level protein structure with a language model

Structure-Infused Protein Language Models