Multi-view biomedical foundation models for molecule-target and property prediction

2024-10-25

arXiv.org

AI Summary

To address insufficient accuracy in molecular target and property prediction for novel drug discovery, this work introduces BioMultiView—the first biomedical multimodal foundation model—integrating molecular graphs, 2D structural images, and textual representations (SMILES/BERT). We pretrain unimodal encoders on a dataset of 200 million molecules and employ dynamic weighted alignment to achieve complementary multimodal feature fusion. Innovatively scaling multiview modeling to the GPCR superfamily (>100 targets), we conduct the first systematic screening of 33 Alzheimer’s disease–associated GPCR targets and identify high-affinity binders. BioMultiView achieves state-of-the-art performance across 18 molecular property and target-binding prediction tasks. Structural modeling validates multiple high-confidence binders and reveals critical binding motifs, substantially enhancing early-stage drug screening efficiency and interpretability.

Technology Category

• Machine Learning: Multi-instance/Multi-view Learning
• Computer Vision: Multi-modal Vision
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Foundation models applied to bio-molecular space hold promise to accelerate drug discovery. Molecular representation is key to building such models. Previous works have typically focused on a single representation or view of the molecules. Here, we develop a multi-view foundation model approach, that integrates molecular views of graph, image and text. Single-view foundation models are each pre-trained on a dataset of up to 200M molecules and then aggregated into combined representations. Our multi-view model is validated on a diverse set of 18 tasks, encompassing ligand-protein binding, molecular solubility, metabolism and toxicity. We show that the multi-view models perform robustly and are able to balance the strengths and weaknesses of specific views. We then apply this model to screen compounds against a large (>100 targets) set of G Protein-Coupled receptors (GPCRs). From this library of targets, we identify 33 that are related to Alzheimer's disease. On this subset, we employ our model to identify strong binders, which are validated through structure-based modeling and identification of key binding motifs.

Problem

• Multimodal Biomedical Modeling
• Drug Discovery Acceleration
• Alzheimer's Disease Treatment

Innovation

• Multi-representation Learning
• Drug Property Prediction
• GPCR Binding Affinity