Ibrahim Muhammad Tukur, Uzairu Adamu, Uba Sani, Shallangwa Gideon Adamu
Department of Chemistry, Faculty of Physical Science, Ahmadu Bello University, P.M.B 1045, Zaria, Kaduna State, Nigeria.
Heliyon. 2020 Feb 7;6(2):e03289. doi: 10.1016/j.heliyon.2020.e03289. eCollection 2020 Feb.
QSAR modelling on Thirty (34) novel quinazoline derivatives (EGFR inhibitors) as non-small cell lung cancer (NSCLC) agents was performed to develop a model with good predictive power that can predict the activities of newly designed compounds that have not been synthesised. The EGFR inhibitors were optimized at B3LYP/6-31G* level of theory using Density Functional Theory (DFT) method. Multi-Linear Regression using Genetic Function Approximation (GFA) method was adopted in building the models. The best one among the models built was selected and reported because it was found to have passed the minimum requirement for the assessment of QSAR models with the following assessment parameters: R of 0.965901, R of 0.893733, Q of 0.940744, R of 0.818991 and LOF of 0.076739. The high predicted power, reliability, robustness of the reported model was verified further by subjecting it to other assessments such VIF, Y-scrambling test and applicability domain. Molecular docking was also employed to elucidate the binding mode of some selected EGFR inhibitors against EGFR receptor () and found that molecule 17 have the highest binding affinity of -9.5 kcal/mol. It was observed that the ligand interacted with the receptor via hydrogen bond, hydrophobic bond, halogen bond, electrostatic bond and others which might me the reason why it has the highest binding affinity. Also, the ADME properties of these selected molecules were predicted and only one molecule (34) was found not orally bioavailable because it violated more than the permissible limit set by Lipinski's rule of five filters. This findings proposed a guidance for designing new potents EGFR inhibitors against their target enzyme.
对34种新型喹唑啉衍生物（表皮生长因子受体抑制剂）作为非小细胞肺癌药物进行了定量构效关系建模，以开发一个具有良好预测能力的模型，该模型可以预测尚未合成的新设计化合物的活性。使用密度泛函理论（DFT）方法在B3LYP/6-31G*理论水平上对表皮生长因子受体抑制剂进行了优化。采用遗传函数逼近（GFA）方法进行多元线性回归来构建模型。在所构建的模型中选择并报告了最佳模型，因为发现它通过了定量构效关系模型评估的最低要求，评估参数如下：R为0.965901，R²为0.893733，Q²为0.940744，R²pred为0.818991，LOF为0.076739。通过对该模型进行其他评估，如方差膨胀因子（VIF）、Y-随机化检验和适用域分析，进一步验证了所报告模型的高预测能力、可靠性和稳健性。还采用分子对接来阐明一些选定的表皮生长因子受体抑制剂与表皮生长因子受体（）的结合模式，发现分子17具有最高的结合亲和力，为-9.5千卡/摩尔。观察到配体通过氢键、疏水键、卤键、静电键等与受体相互作用，这可能是其具有最高结合亲和力的原因。此外，预测了这些选定分子的药物代谢动力学性质，发现只有一个分子（34）不具有口服生物利用度，因为它违反了比Lipinski五规则过滤器设定的允许限度更多的规则。该发现为设计针对其靶酶的新型强效表皮生长因子受体抑制剂提供了指导。
