[2510.02578] FLOWR.root: A flow matching based foundation model for joint multi-purpose structure-aware 3D ligand generation and affinity prediction

Quantitative Biology > Biomolecules arXiv:2510.02578 (q-bio) [Submitted on 2 Oct 2025 (v1), last revised 3 Mar 2026 (this version, v4)] Title:FLOWR.root: A flow matching based foundation model for joint multi-purpose structure-aware 3D ligand generation and affinity prediction Authors:Julian Cremer, Tuan Le, Mohammad M. Ghahremanpour, Emilia Sługocka, Filipe Menezes, Djork-Arné Clevert View a PDF of the paper titled FLOWR.root: A flow matching based foundation model for joint multi-purpose structure-aware 3D ligand generation and affinity prediction, by Julian Cremer and 5 other authors View PDF Abstract:We present this http URL, an SE(3)-equivariant flow-matching model for pocket-aware 3D ligand generation with joint potency and binding affinity prediction and confidence estimation. The model supports de novo generation, interaction- and pharmacophore-conditional sampling, fragment elaboration and replacement, and multi-endpoint affinity prediction (pIC50, pKi, pKd, pEC50). Training combines large-scale ligand libraries with mixed-fidelity protein-ligand complexes, refined on curated co-crystal datasets and adapted to project-specific data through parameter-efficient finetuning. The base this http URL model achieves state-of-the-art performance in unconditional 3D molecule and pocket-conditional ligand generation. On HiQBind, the pre-trained and finetuned model demonstrates highly accurate affinity predictions, and outperforms recent state-of-the-art methods such as Boltz...