Symmetry-Aware Flow Matching for End-to-end Molecular Crystal Generation
Abstract
Molecular crystal packing shapes properties from drug bioavailability to charge transport, yet generating realistic structures requires coordinating molecular flexibility, intermolecular interactions and symmetry. Here we introduce SALA, a flow-matching model built on state--context separation: it evolves only asymmetric-unit (ASU) atomic coordinates and symmetry-compatible lattice parameters while predicting their updates from the full periodic environment. Conditioned on molecular graphs and a space group, SALA generates molecular conformation, lattice geometry and packing end to end while preserving the specified symmetry. Across 799 held-out crystals spanning four chemical classes and all seven crystal systems, SALA achieves a 72.0\% hit rate from 50 candidates per target, compared with 8.1\% for an atomistic full-cell generative baseline. Mean best-candidate packing root-mean-square deviation decreases from 4.48 to 1.70~Å, with the advantage widening as system size and symmetry multiplicity increase. SALA thus enables end-to-end generation of complex, symmetry-constrained molecular crystals, with potential applications in CSP, structure-based pharmaceutical polymorph design and organic molecular materials discovery.
Related
- Symmetry-aware Bayesian flow networks for crystal generation
- Curvature-Aware Flow Matching for Molecular Structure Generation
- An unlearnable target: diagnosing the orientation gauge in PXRD-conditioned molecular-crystal flow matching
- FLOWR.ROOT – A flow matching-based foundation model for joint multi-purpose structure-aware 3D ligand generation and affinity prediction
- FLOWR: flow matching for structure-aware de novo, interaction- and fragment-based ligand generation
BibTeX
@misc{cai2026symmetryaware,
title = {{Symmetry-Aware Flow Matching for End-to-end Molecular Crystal Generation}},
author = {Wendi Cai and Fanyang Mo},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.34690},
eprint = {2609.34690},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.34690},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - UNPB
TI - Symmetry-Aware Flow Matching for End-to-end Molecular Crystal Generation
AU - Wendi Cai
AU - Fanyang Mo
PY - 2026
DA - 2026/09/28
PB - arXiv
DO - 10.48550/arxiv.2609.34690
UR - https://arxiv.org/abs/2609.34690
AB - Molecular crystal packing shapes properties from drug bioavailability to charge transport, yet generating realistic structures requires coordinating molecular flexibility, intermolecular interactions and symmetry. Here we introduce SALA, a flow-matching model built on state--context separation: it evolves only asymmetric-unit (ASU) atomic coordinates and symmetry-compatible lattice parameters while predicting their updates from the full periodic environment. Conditioned on molecular graphs and a space group, SALA generates molecular conformation, lattice geometry and packing end to end while preserving the specified symmetry. Across 799 held-out crystals spanning four chemical classes and all seven crystal systems, SALA achieves a 72.0\% hit rate from 50 candidates per target, compared with 8.1\% for an atomistic full-cell generative baseline. Mean best-candidate packing root-mean-square deviation decreases from 4.48 to 1.70~Å, with the advantage widening as system size and symmetry multiplicity increase. SALA thus enables end-to-end generation of complex, symmetry-constrained molecular crystals, with potential applications in CSP, structure-based pharmaceutical polymorph design and organic molecular materials discovery.
ER -