arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.07351cond-mat.mtrl-sci
收藏

AtomMOF:基于原子流匹配的MOF-吸附质结构预测

AtomMOF: All-Atom Flow Matching for MOF-Adsorbate Structure Prediction

Korea Advanced Institute of Science and Technology(韩国科学技术院)

Nayoung Kim, Honghui Kim, Sihyun Yu, Minkyu Kim, Seongsu Kim, Sungsoo Ahn

AI总结:

AtomMOF通过原子流匹配方法提升MOF-吸附质结构预测的准确性和效率

摘要:

深度生成模型在建模金属-有机框架(MOF)方面显示出潜力,但现有方法(1)依赖于粗粒表示,假设固定键长和键角,(2)忽视MOF-吸附质相互作用,这对下游应用至关重要。我们引入AtomMOF,一种可扩展的基于流的模型,基于一个全部原子扩散变换器,将构建块和吸附质的2D分子图直接映射到平衡的3D结构,而无需施加结构约束。我们进一步提出了多孔晶体生成的扩展定律,表明随着模型容量的增加,性能有可预测的提升,并引入了由机器学习互原子势引导的费曼-卡茨引导,以提高几何有效性和采样稳定性。在(仅MOF)BW数据集上,AtomMOF将匹配率提高了35.00%,并将RMSD降低了32.64%。在ODAC25数据集(MOF-吸附质)上,AtomMOF在恢复吸附配置方面比广义canonical蒙特卡洛方法更高效,并能识别出比参考数据集更低吸附能的候选物。代码可在https://github.com/nayoung10/AtomMOF上获得。

英文摘要:

Metal-organic frameworks (MOFs) are promising materials for applications such as direct air capture, where performance depends on how adsorbates bind within the framework. Accurately predicting MOF-adsorbate configurations is therefore important for screening candidate materials. However, traditional approaches are computationally expensive and require a known host structure, while existing generative models rely on rigid-body assumptions and do not explicitly model adsorbates. We introduce AtomMOF, a scalable all-atom flow-matching model that jointly predicts MOF-adsorbate structures from discrete building blocks and adsorbates. Built on a Diffusion Transformer with a building block-based pairwise attention bias, AtomMOF operates in an unconstrained all-atom space and exhibits clear scaling behavior. To improve the structural validity of flexible all-atom models, we also propose Feynman-Kac (FK) steering with machine-learned interatomic potentials (MLIPs). On the BW dataset, AtomMOF achieves a 85.61% relative increase in match rate and a 34.32% relative reduction in RMSD compared with prior work. MLIP-guided steering further improves validity by 28.7% and reduces formation energy error by 86.5%. On ODAC25, AtomMOF generates adsorption configurations faster than GCMC and, when combined with MLIP relaxation, identifies lower-energy configurations than those in the reference dataset.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑