FLOWR.root: A flow matching based foundation model for joint multi-purpose structure-aware 3D ligand generation and affinity prediction

2025-10-02

AI Summary

Existing 3D ligand generation models suffer from geometric distortion, high strain energy, and a trade-off between generation quality and binding affinity prediction accuracy. To address these challenges, we propose FLOWR.root—the first foundational model integrating equivariance and flow matching to jointly perform pocket-conditioned 3D ligand generation and multi-endpoint affinity prediction (pIC50/pKi/pKd/pEC50). Our method introduces importance sampling for inference-time structural refinement, hybrid-fidelity training, parameter-efficient fine-tuning, and a quantum-mechanics-validated geometric constraint framework. FLOWR.root achieves state-of-the-art performance in both unconditional and pocket-conditioned generation tasks. On the SPINDR and Schrödinger FEP+/OpenFE benchmarks, it significantly outperforms prior methods in affinity prediction—delivering higher computational efficiency and stronger correlation with experimental measurements. The framework comprehensively supports key drug discovery workflows, including de novo design, pharmacophore-guided optimization, and fragment growth.

Technology Category

• Machine Learning: Other Foundations of Machine Learning
• Search and Optimization: Sampling/Simulation-based Search
• Natural Language Processing: Generation

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

We present Flowr.root, an equivariant flow-matching model for pocket-aware 3D ligand generation with joint binding affinity prediction and confidence estimation. The model supports de novo generation, pharmacophore-conditional sampling, fragment elaboration, and multi-endpoint affinity prediction (pIC50, pKi, pKd, pEC50). Training combines large-scale ligand libraries with mixed-fidelity protein-ligand complexes, followed by refinement on curated co-crystal datasets and parameter-efficient finetuning for project-specific adaptation. Flowr.root achieves state-of-the-art performance in unconditional 3D molecule generation and pocket-conditional ligand design, producing geometrically realistic, low-strain structures. The integrated affinity prediction module demonstrates superior accuracy on the SPINDR test set and outperforms recent models on the Schrodinger FEP+/OpenFE benchmark with substantial speed advantages. As a foundation model, Flowr.root requires finetuning on project-specific datasets to account for unseen structure-activity landscapes, yielding strong correlation with experimental data. Joint generation and affinity prediction enable inference-time scaling through importance sampling, steering molecular design toward higher-affinity compounds. Case studies validate this: selective CK2alpha ligand generation against CLK3 shows significant correlation between predicted and quantum-mechanical binding energies, while ERalpha and TYK2 scaffold elaboration demonstrates strong agreement with QM calculations. By integrating structure-aware generation, affinity estimation, and property-guided sampling, Flowr.root provides a comprehensive foundation for structure-based drug design spanning hit identification through lead optimization.

Problem

• Develops joint 3D ligand generation and affinity prediction model
• Enables structure-aware drug design from hit identification to optimization
• Integrates property-guided sampling for higher-affinity compound generation

Innovation

• Equivariant flow-matching model for pocket-aware 3D ligand generation
• Joint binding affinity prediction with confidence estimation
• Parameter-efficient finetuning for project-specific adaptation