Zubatyuk Roman, Biczysko Malgorzata, Ranasinghe Kavindri, Moriarty Nigel W, Gokcan Hatice, Kruse Holger, Poon Billy K, Adams Paul D, Waller Mark P, Roitberg Adrian E, Isayev Olexandr, Afonine Pavel V
Department of Chemistry, Carnegie Mellon University, Pittsburgh, PA 15213, USA.
Faculty of Chemistry, University of Wrocław, F. Joliot-Curie 14, 50-383 Wrocław, Poland.
bioRxiv. 2024 Jul 21:2024.07.21.604493. doi: 10.1101/2024.07.21.604493.
Cryo-EM and X-ray crystallography provide crucial experimental data for obtaining atomic-detail models of biomacromolecules. Refining these models relies on library-based stereochemical restraints, which, in addition to being limited to known chemical entities, do not include meaningful noncovalent interactions relying solely on nonbonded repulsions. Quantum mechanical (QM) calculations could alleviate these issues but are too expensive for large molecules. We present a novel AI-enabled Quantum Refinement (AQuaRef) based on AIMNet2 neural network potential mimicking QM at substantially lower computational costs. By refining 41 cryo-EM and 30 X-ray structures, we show that this approach yields atomic models with superior geometric quality compared to standard techniques, while maintaining an equal or better fit to experimental data.
冷冻电镜和X射线晶体学为获取生物大分子的原子细节模型提供了关键的实验数据。完善这些模型依赖于基于库的立体化学约束，这些约束除了限于已知的化学实体外，不包括仅依赖非键排斥的有意义的非共价相互作用。量子力学（QM）计算可以缓解这些问题，但对于大分子来说成本太高。我们提出了一种基于AIMNet2神经网络势的新型人工智能量子精修（AQuaRef）方法，以低得多的计算成本模拟QM。通过对41个冷冻电镜结构和30个X射线结构进行精修，我们表明，与标准技术相比，这种方法产生的原子模型具有更好的几何质量，同时与实验数据的拟合度相同或更好。
