Structure-Aware Fusion with Progressive Injection for Multimodal Molecular Representation Learning

Yan Sun (New Jersey Institute of Technology) · Pingzhao Hu (Western University) · Zihao Jing (University of Western Ontario) · Yan Yi Li (University of Toronto) · Sugitha Janarthanan (Western University) · Alana Deng (University of Western Ontario)

3d conformerbenchmark taskscross-modal enrichmentinformation propagation long-range dependency modeling modality collapsemolecular representationmoleculenetmultimodal fusionprogressive injectionrobustness to noisesequence streamstate space backbonestructured fusion pipelinetherapeutics data commons

Multimodal molecular models often suffer from 3D conformer unreliability and modality collapse, limiting their robustness and generalization. We propose MuMo, a structured multimodal fusion framework that addresses these challenges in molecular representation through two key strategies. To reduce the instability of conformer-dependent fusion, we design a Structured Fusion Pipeline (SFP) that combines 2D topology and 3D geometry into a unified and stable structural prior. To mitigate modality collapse caused by naive fusion, we introduce a Progressive Injection (PI) mechanism that asymmetrically integrates this prior into the sequence stream, preserving modality-specific modeling while enabling cross-modal enrichment. Built on a state space backbone, MuMo supports long-range dependency modeling and robust information propagation. Across 29 benchmark tasks from Therapeutics Data Commons (TDC) and MoleculeNet, MuMo achieves an average improvement of 2.7% over the best-performing baseline on each task, ranking first on 22 of them, including a 27% improvement on the LD50 task. These results validate its robustness to 3D conformer noise and the effectiveness of multimodal fusion in molecular representation. The code is available at: github.com/selmiss/MuMo.