Alikhani Radin, Ebadi Ahmad, Karami Pari, Shahbipour Sara, Razzaghi-Asl Nima
Students Research Committee, School of Pharmacy, Ardabil University of Medical Sciences, Ardabil, Iran.
Department of Medicinal Chemistry, School of Pharmacy, Hamadan University of Medical Sciences, Hamadan, Iran.
Iran J Pharm Res. 2021 Summer;20(3):560-576. doi: 10.22037/ijpr.2020.113644.14409.
Computer-aided drug design provides broad structural modifications to evolving bioactive molecules without an immediate requirement to observe synthetic restraints or tedious protocols. Subsequently, the most promising guidelines with regard to synthetic and biological resources may be focused on upcoming steps. Molecular docking is common drug design techniques since it predicts ligand-receptor interaction modes and associated binding affinities. Current docking simulations suffer serious constraints in estimating accurate ligand-receptor binding affinities despite several advantages and historical results. Response surface method (RSM) is an efficient statistical approach for modeling and optimization of various pharmaceutical systems. With the aim of unveiling the full potential of RSM in optimizing molecular docking simulations, this study particularly focused on binding affinity prediction of citalopram-serotonin transporter (SERT) and donepezil-acetyl cholinesterase (AChE) complexes. For this purpose, Box-Behnken design of experiments (DOE) was used to develop a trial matrix for simultaneous variations of AutoDock4.2 driven binding affinity data with selected factor levels. Responses of all docking trials were considered as estimated protein inhibition constants with regard to validated data for each drug. The output matrix was subjected to statistical analysis and constructing polynomial quadratic models. Numerical optimization steps to attain ideal docking accuracies revealed that more accurate results might be envisaged through the best combination of factor levels and considering factor interactions. Results of the current study indicated that the application of RSM in molecular docking simulations might lead to optimized docking protocols with more stable estimates of ligand-target interactions and hence better correlation of data.
计算机辅助药物设计为不断演变的生物活性分子提供了广泛的结构修饰，而无需立即考虑合成限制或繁琐的实验方案。随后，关于合成和生物资源的最有前景的指导方针可能集中在后续步骤上。分子对接是常用的药物设计技术，因为它可以预测配体 - 受体相互作用模式及相关的结合亲和力。尽管当前的对接模拟有诸多优点和既往成果，但在估计准确的配体 - 受体结合亲和力方面仍存在严重限制。响应面法（RSM）是一种用于各种药物系统建模和优化的有效统计方法。为了揭示RSM在优化分子对接模拟中的全部潜力，本研究特别聚焦于西酞普兰 - 血清素转运体（SERT）和多奈哌齐 - 乙酰胆碱酯酶（AChE）复合物的结合亲和力预测。为此，采用Box - Behnken实验设计（DOE）来开发一个试验矩阵，用于同时改变AutoDock4.2驱动的结合亲和力数据及选定的因子水平。所有对接试验的响应被视为针对每种药物的验证数据的估计蛋白质抑制常数。对输出矩阵进行统计分析并构建多项式二次模型。实现理想对接精度的数值优化步骤表明，通过因子水平的最佳组合并考虑因子相互作用，可以设想出更准确的结果。本研究结果表明，RSM在分子对接模拟中的应用可能会产生优化的对接方案，对配体 - 靶点相互作用有更稳定的估计，从而使数据具有更好的相关性。
