Yang Zhuo, Liu Yingtao, Chen Zhaoqiang, Xu Zhijian, Shi Jiye, Chen Kaixian, Zhu Weiliang
Drug Discovery and Design Center, Key Laboratory of Receptor Research, Shanghai Institute of Materia Medica, Chinese Academy of Sciences, 555 Zuchongzhi Road, Shanghai, 201203, China,
J Mol Model. 2015 Jun;21(6):138. doi: 10.1007/s00894-015-2681-6. Epub 2015 May 10.
A quantum mechanics-based scoring function for halogen bonding interaction, namely XBScore(QM), is developed based on 18,135 sets of geometrical and energetical parameters optimized at M06-2X/aug-cc-pVDZ level. Applying the function on typical halogen bonding systems from Protein Data Bank demonstrates its strong ability of predicting halogen bonding as attractive interaction with strength up to -4 kcal mol(-1). With a diverse set of proteins complexed with halogenated ligands, a systematic evaluation demonstrates the integrative advantage of XBScore(QM) over 12 other scoring functions on halogen bonding in four aspects, viz. pseudo docking power, ranking power, scoring power, and genuine docking power. Thus, this study not only provides a practicable scoring function of halogen bonding for high throughput virtual screening, but also serves as a benchmark for evaluating the performance of current scoring functions on characterizing halogen bonding.
基于量子力学开发了一种用于卤键相互作用的评分函数,即XBScore(QM),它基于在M06-2X/aug-cc-pVDZ水平上优化的18135组几何和能量参数。将该函数应用于蛋白质数据库中的典型卤键系统,证明了其强大的预测卤键作为吸引相互作用的能力,强度可达-4 kcal mol(-1)。通过对一系列与卤化配体复合的蛋白质进行系统评估,结果表明XBScore(QM)在卤键方面比其他12种评分函数在四个方面具有综合优势,即伪对接能力、排序能力、评分能力和真正的对接能力。因此,本研究不仅为高通量虚拟筛选提供了一种实用的卤键评分函数,也为评估当前评分函数在表征卤键方面的性能提供了一个基准。