Triangle Multiplication Is All You Need For Biomolecular Structure Representations

TL;DR

Pairmixer is a novel protein structure prediction model that replaces triangle attention with a more efficient approach, significantly reducing computational costs while maintaining high accuracy, enabling large-scale applications.

Contribution

The paper introduces Pairmixer, a new method that simplifies geometric reasoning in protein modeling, improving efficiency over existing models like Pairformer.

Findings

Up to 4x faster inference on long sequences

Reduces training cost by 34%

Scales to sequences 30% longer than previous models

Abstract

AlphaFold has transformed protein structure prediction, but emerging applications such as virtual ligand screening, proteome-wide folding, and de novo binder design demand predictions at a massive scale, where runtime and memory costs become prohibitive. A major bottleneck lies in the Pairformer backbone of AlphaFold3-style models, which relies on computationally expensive triangular primitives-especially triangle attention-for pairwise reasoning. We introduce Pairmixer, a streamlined alternative that eliminates triangle attention while preserving higher-order geometric reasoning capabilities that are critical for structure prediction. Pairmixer substantially improves computational efficiency, matching state-of-the-art structure predictors across folding and docking benchmarks, delivering up to 4x faster inference on long sequences while reducing training cost by 34%. Its efficiency alleviates the computational burden of downstream applications such as modeling large protein complexes, high-throughput ligand and binder screening, and hallucination-based design. Within BoltzDesign, for example, Pairmixer delivers over 2x faster sampling and scales to sequences ~30% longer than the memory limits of Pairformer. Code is available at https://github.com/genesistherapeutics/pairmixer.