DeepMoLM: Leveraging Visual and Geometric Structural Information for Molecule-Text Modeling
This addresses the challenge of accurately modeling molecules from images for drug discovery and chemical literature mining, representing an incremental improvement over existing methods by integrating geometric invariants.
The paper tackles the problem of interpreting molecular images and generating outputs consistent with 3D geometry in drug discovery by proposing DeepMoLM, a dual-view framework that fuses visual and geometric information, resulting in a 12.3% relative METEOR gain over generalist baselines in PubChem captioning and competitive performance on property queries and description generation.
AI models for drug discovery and chemical literature mining must interpret molecular images and generate outputs consistent with 3D geometry and stereochemistry. Most molecular language models rely on strings or graphs, while vision-language models often miss stereochemical details and struggle to map continuous 3D structures into discrete tokens. We propose DeepMoLM: Deep Molecular Language M odeling, a dual-view framework that grounds high-resolution molecular images in geometric invariants derived from molecular conformations. DeepMoLM preserves high-frequency evidence from 1024 $\times$ 1024 inputs, encodes conformer neighborhoods as discrete Extended 3-Dimensional Fingerprints, and fuses visual and geometric streams with cross-attention, enabling physically grounded generation without atom coordinates. DeepMoLM improves PubChem captioning with a 12.3% relative METEOR gain over the strongest generalist baseline while staying competitive with specialist methods. It produces valid numeric outputs for all property queries and attains MAE 13.64 g/mol on Molecular Weight and 37.89 on Complexity in the specialist setting. On ChEBI-20 description generation from images, it exceeds generalist baselines and matches state-of-the-art vision-language models. Code is available at https://github.com/1anj/DeepMoLM.