DrugBLIP: exploring the protein-molecule interaction mechanisms with a multi-task learning graph transformer

<h4>Motivation</h4>Traditional drug discovery methods are costly and inefficient, while existing deep learning approaches remain limited by task specificity and practical applicability. Accurately modeling protein-molecule interactions is critical for advancing virtual screening, docking, and drug design.<h4>Results</h4>We propose DrugBLIP, a multi-task graph transformer model based on SE(3)-equivariant architectures, to unify protein-molecule interaction learning. By integrating contrastive learning, matching tasks, and docking optimization, DrugBLIP captures 3D spatial relationships through a hybrid graph transformer framework. Evaluations demonstrate state-of-the-art performance: DrugBLIP achieves an AUROC of 0.8217 and BEDROC of 0.5743 on virtual screening, outperforming traditional and deep learning baselines by 10%-127% across metrics. It also attains 91.2% top-1 docking success on CASF-2016 and 41.8% target fishing accuracy, showcasing robustness in diverse scenarios. Additionally, DrugBLIP reduces computational time by 700× compared to traditional docking tools.<h4>Availability and implementation</h4>Code is available at https://github.com/Wolkenwandler/DrugBLIP and archived at Zenodo with DOI: 10.5281/zenodo.16990700.