对称感知流匹配用于端到端分子晶体生成
Symmetry-Aware Flow Matching for End-to-end Molecular Crystal Generation
- Peking University Shenzhen Graduate School（北京大学深圳研究生院）
- Peking University（北京大学）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
本文提出对称感知流匹配模型SALA，通过状态-上下文分离生成分子晶体，在799个晶体上实现72%命中率，显著优于基线，适用于晶体结构预测和药物设计。
AI中文摘要：
分子晶体堆积决定了从药物生物利用度到电荷传输等性质，然而生成真实结构需要协调分子柔性、分子间相互作用和对称性。这里我们介绍SALA，一种基于状态-上下文分离构建的流匹配模型：它仅演化不对称单元（ASU）原子坐标和对称兼容的晶格参数，同时从完整周期环境中预测其更新。在分子图和空间群条件下，SALA端到端生成分子构象、晶格几何和堆积，同时保持指定对称性。在跨越四个化学类别和所有七个晶系的799个保留晶体中，SALA从每个目标50个候选中实现了72.0%的命中率，而原子级全细胞生成基线为8.1%。最佳候选堆积均方根偏差从4.48 Å降至1.70 Å，随着系统尺寸和对称多重性增加，优势扩大。因此，SALA能够端到端生成复杂、对称约束的分子晶体，在CSP、基于结构的药物多晶型设计和有机分子材料发现中具有潜在应用。
英文摘要：
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.
