Thin Bridges For Drug Text Alignment: Lightweight Contrastive Learning For Target Specific Drug Retrieval

Abstract

Multimodal foundation models hold promise for drug discovery and biomedical applications, but most existing approaches rely on heavy pretraining or large scale multimodal corpora. We investigate whether thin contrastive bridges, lightweight projection heads over frozen unimodal encoders can align chemical and textual representations without training a full multimodal model. Using paired mechanisms from ChEMBL, we align ECFP4 molecular fingerprints with biomedical sentence embeddings through dual linear projections trained with a contrastive objective. To better handle drugs sharing the same therapeutic target, we incorporate hard negative weighting and a margin loss. Evaluation under scaffold based splits, which require generalization across disjoint chemical cores, demonstrates that our approach achieves non-trivial cross modal alignment and substantially improves within target discrimination compared to frozen baselines. These results suggest that thin bridges offer a compute efficie

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keytupakula2025thin

Related papers

• Towards Cross-modal Text-molecule Retrieval With Better Modality Alignment (2024)4.52
• Geomclip: Contrastive Geometry-text Pre-training For Molecules (2024)1.40
• Weakly Supervised Cross-modal Learning In High-content Screening (2023)2.26
• BRIDGE: Multimodal-to-text Retrieval Via Reinforcement-learned Query Alignment (2026)0.00
• Textme: Bridging Unseen Modalities Through Text Descriptions (2026)0.00
• Biobridge: Bridging Biomedical Foundation Models Via Knowledge Graphs (2023)0.00
• Multi-task Cross-modal Learning For Chest X-ray Image Retrieval (2026)0.00
• Benchmarking Vision-language Contrastive Methods For Medical Representation Learning (2024)0.00