Learning to Align Molecules and Proteins: A Geometry-Aware Approach to Binding Affinity
Preprint
Abstract
Accurate prediction of drug-target binding affinity can accelerate drug discovery by prioritizing promising compounds before costly wet-lab screening. While deep learning has advanced this task, most models fuse ligand and protein representations via simple concatenation and lack explicit geometric regularization, resulting in poor generalization across chemical space and time. We introduce FIRM-DTI, a lightweight framework that conditions molecular embeddings on protein embeddings through a feature-wise linear modulation (FiLM) layer and enforces metric structure with a triplet loss. An RBF regression head operating on embedding distances yields smooth, interpretable affinity predictions. Despite its modest size, FIRM-DTI achieves state-of-the-art performance on the Therapeutics Data Commons DTI-DG benchmark, as demonstrated by an extensive ablation study and out-of-domain evaluation. Our results underscore the value of conditioning and metric learning for robust drug-target affinity prediction.Keywords
All Related Versions
This publication has 9 references indexed in Scilit:
- Aggregating residue-level protein language model embeddings with optimal transportBioinformatics Advances, 2024
- MolE: a foundation model for molecular graphs using disentangled attentionNature Communications, 2024
- Contrastive learning in protein language space predicts interactions between drugs and protein targetsProceedings of the National Academy of Sciences, 2023
- Evolutionary-scale prediction of atomic-level protein structure with a language modelScience, 2023
- MolTrans: Molecular Interaction Transformer for drug–target interaction predictionBioinformatics, 2020
- DeepConv-DTI: Prediction of drug-target interactions via deep learning with convolution on protein sequencesPLOS Computational Biology, 2019
- Compound–protein interaction prediction with end-to-end learning of neural networks for graphs and sequencesBioinformatics, 2018
- Comprehensive analysis of kinase inhibitor selectivityNature Biotechnology, 2011
- BindingDB: a web-accessible database of experimentally determined protein-ligand binding affinitiesNucleic Acids Research, 2006