Host-guest binding is a challenging problem in computer simulation. The prediction of binding affinities between hosts and guests is an important part of the statistical assessment of the modeling of proteins and ligands (SAMPL) challenges. In this work, the volume-based variant of well-tempered metadynamics is employed to calculate the binding affinities of the host-guest systems in the SAMPL6 challenge. By biasing the spherical coordinates describing the relative position of the host and the guest, the initial-configuration-induced bias vanishes and all possible binding poses are explored. The agreement between the predictions and the experimental results and the observation of new binding poses indicate that the volume-based technique serves as a nice candidate for the calculation of binding free energies and the search of the binding poses.

中文翻译：

---

SAMPL6 主客体绑定亲和力和绑定姿势来自球坐标偏向模拟。

主客绑定是计算机模拟中的一个具有挑战性的问题。宿主和客体之间结合亲和力的预测是蛋白质和配体建模 (SAMPL) 挑战的统计评估的重要组成部分。在这项工作中，采用基于体积的温和元动力学变体来计算 SAMPL6 挑战中主客体系统的结合亲和力。通过偏置描述主客体相对位置的球坐标，初始配置引起的偏差消失，并探索所有可能的结合姿势。预测和实验结果之间的一致性以及对新结合位姿的观察表明，基于体积的技术是计算结合自由能和搜索结合位姿的一个很好的候选者。