CASTER-DTA: Equivariant Graph Neural Networks for Predicting Drug-Target Affinity

Accurately determining the binding affinity of a ligand with a protein is important for drug design, development, and screening. With the advent of accessible protein structure prediction methods such as AlphaFold, predicted protein 3D structures are readily available; however, methods for predicting binding affinity currently do not take full advantage of 3D protein information. Here, we present CASTER-DTA (Cross-Attention with Structural Target Equivariant Representations for Drug-Target Affinity), which uses an equivariant graph neural network to learn more robust protein representations alongside a standard graph neural network to learn molecular representations to predict drug-target affinity. We augment these representations by incorporating an attention-based mechanism between protein residues and drug atoms to improve interpretability. We show that CASTER-DTA represents a state-of-the-art improvement on multiple benchmarks for predicting drug-target affinity and that it generates novel insights for several related tasks. We then apply CASTER-DTA to create a large resource of the binding affinities of every FDA-approved drug against every protein in the human proteome and make these predictions freely available for download. We also make available a web server for researchers to apply a pretrained CASTER-DTA model for predicting binding affinities between arbitrary proteins and drugs. CASTER-DTA is a novel architecture that makes use of equivariant graph neural networks to predict drug-target affinity, enabling rapid and interpretable predictions about how proteins and molecules bind to each other. Equivariant graph neural networks like those used in CASTER-DTA allow for more robust usage of protein 3D structural information and improve performance over non-equivariant approaches. CASTER-DTA includes a cross-attention layer to update protein residue and molecule atom embeddings that allows for interpretability about which residues are involved in the prediction. We can use this architecture for a variety of downstream tasks, and we have used CASTER-DTA to generate a resource of predicted binding affinities for every FDA-approved drug against every protein in the human proteome.