TRIDENT: Tri-Modal Molecular Representation Learning with Taxonomic Annotations and Local Correspondence
This addresses the problem of limited multimodal learning in molecular science by incorporating overlooked textual and taxonomic information, though it is incremental as it builds on existing multimodal paradigms.
The paper tackles molecular property prediction by introducing TRIDENT, a framework that integrates SMILES, textual descriptions, and taxonomic annotations to learn rich representations, achieving state-of-the-art performance on 11 downstream tasks.
Molecular property prediction aims to learn representations that map chemical structures to functional properties. While multimodal learning has emerged as a powerful paradigm to learn molecular representations, prior works have largely overlooked textual and taxonomic information of molecules for representation learning. We introduce TRIDENT, a novel framework that integrates molecular SMILES, textual descriptions, and taxonomic functional annotations to learn rich molecular representations. To achieve this, we curate a comprehensive dataset of molecule-text pairs with structured, multi-level functional annotations. Instead of relying on conventional contrastive loss, TRIDENT employs a volume-based alignment objective to jointly align tri-modal features at the global level, enabling soft, geometry-aware alignment across modalities. Additionally, TRIDENT introduces a novel local alignment objective that captures detailed relationships between molecular substructures and their corresponding sub-textual descriptions. A momentum-based mechanism dynamically balances global and local alignment, enabling the model to learn both broad functional semantics and fine-grained structure-function mappings. TRIDENT achieves state-of-the-art performance on 11 downstream tasks, demonstrating the value of combining SMILES, textual, and taxonomic functional annotations for molecular property prediction.