Language-Informed Flow Matching for Trend-Guided Structure-Based 3D Molecular Generation
2026-08-31 • Machine Learning
Machine Learning
AI summaryⓘ
The authors developed LiFT, a new method for designing 3D molecules that fit well with targeted proteins and follow chemical rules. Unlike other methods that need extra tuning or complicated adjustments, LiFT uses language information from chemical formulas (SMILES) to guide molecular shape generation in a smoother way. Their approach combines language and 3D geometry through special neural networks, improving molecule design quality and chemical correctness without extra training. Tests show LiFT works well on benchmark tasks, helping create better drug-like molecules automatically.
Structure-based drug designSMILES3D molecular generationFlow MatchingCross-modal learningSemantic priorsODE integrationScaffold hoppingChemical validityVelocity field modulation
Authors
Tianyu Gao, Zhikai Su, Jiashu Li, Wenjun Gao, Zichuan Ying, Zhe Zhao, Fei Zhang, Ye Wei
Abstract
Structure-based drug design (SBDD) requires ligands that satisfy both 3D target affinity and 1D chemical validity. Existing controllable generation methods often rely on task-specific fine-tuning or externally imposed sampling-time guidance, adding cost and potentially conflicting with evolving 3D geometric constraints. We propose LiFT, a language-informed cross-modal framework built on Flow Matching for trend-guided 3D molecular generation across both de novo design and scaffold hopping. LiFT uses a "Sense-Evolve-Assemble" agent to generate target-aware SMILES as intermediate chemical conditions, from which a pre-trained chemical foundation model extracts continuous semantic priors. These priors are integrated into geometric generation through a lightweight semantic projector with zero-initialized adaptive normalization for stable cross-modal conditioning. We further introduce a Self-Conditioned Decoupled Router (SCDR), which modulates the velocity field according to intermediate structural states during ODE integration. Experiments on Cross-Docked2020 show that LiFT achieves competitive distribution matching while improving medicinal chemistry metrics and maintaining competitive structural validity under task-steering settings without additional generator fine-tuning. Our results suggest that language-derived chemical priors provide effective trend-level guidance for 3D molecular generation. Code and released artifacts are available at https://github.com/kasurl/LiFT.