Philipps Universität Marburg, Institut für Pharmazeutische Chemie, Marbacher Weg 6, Marburg 35037, Germany.
J Med Chem. 2021 Apr 22;64(8):4662-4676. doi: 10.1021/acs.jmedchem.0c02115. Epub 2021 Apr 2.
The consideration of interactions involving water molecules in protein-ligand binding is widely appreciated in drug discovery nowadays. However, it is not ultimately clear how insights about these interactions translate into molecular design concepts. In this work, we introduce a computational strategy that, trained with high-precision experimental data, allows for the decomposition of water-related thermodynamic properties into chemically relevant building blocks (BBs) of a given ligand scaffold. For each of these BBs, a score based on solvation energy and entropy is computed, thus enabling the analysis of solvent-related affinity contributions for individual BBs. We find the nonvariable BB in a congeneric ligand pair to have a larger impact on the binding affinity than the variable part thus suggesting strong cooperative effects. Furthermore, we find enhanced solute-solvent interactions for a BB due to the presence of a C-F bond. Our investigation may be used to design drug molecules with tailored solvent thermodynamic properties.
在当今的药物发现中，人们广泛认识到考虑蛋白质-配体结合中水分子的相互作用。然而，这些相互作用的见解如何转化为分子设计概念尚不完全清楚。在这项工作中，我们引入了一种计算策略，该策略经过高精度实验数据的训练，允许将与水相关的热力学性质分解为给定配体支架的化学相关构建块（BB）。对于每个这样的 BB，都会计算基于溶剂化能和熵的得分，从而能够分析单个 BB 的与溶剂相关的亲和性贡献。我们发现同系物配体对中非变量 BB 对结合亲和力的影响大于变量部分，这表明存在强烈的协同效应。此外，我们发现由于存在 C-F 键，BB 中溶质-溶剂相互作用增强。我们的研究可用于设计具有定制溶剂热力学性质的药物分子。
