Colorectal cancer is common to both sexes; third in terms of morbidity and second in terms of mortality, accounting for 10% and 9.2% of cancer cases in men and women globally. Although drugs such as bevacizumab, Camptosar, and cetuximab are being used to manage colorectal cancer, the efficacy of the drugs has been reported to vary from patient to patient. These drugs have also been reported to have varying degrees of side effects; thus, the need for novel drug therapies with better efficacy and lesser side effects. In silico drugs design methods provide a faster and cost-effect method for lead identification and optimization. The aim of this study, therefore, was to design novel imidazol-5-ones via in silico design methods. A QSAR model was built using the genetic function algorithm method to model the cytotoxicity of the compounds against the HCT116 colorectal cancer cell line. The built model had statistical parameters; R2 = 0.7397, R2adj = 0.6712, Q2cv = 0.5547, and R2ext. = 0.7202 and revealed the cytotoxic activity of the compounds to be dependent on the molecular descriptors nS, GATS5s, VR1_Dze, ETA_dBetaP, and L3i. These molecular descriptors were poorly correlated (VIF < 4.0) and made unique contributions to the built model. The model was used to design a novel set of derivatives via the ligand-based drug design approach. Compounds e, h, j, and l showed significantly better cytotoxicity (IC50 < 5.0 μM) compared to the template. The interaction of the compounds with the CDK2 enzyme (PDB ID: 6GUE) was investigated via molecular docking study. The compounds were potent inhibitors of the enzyme having binding affinity of range −10.8 to −11.0 kcal/mol and primarily formed hydrogen bond interaction with lysine, aspartic acid, leucine, and histidine amino acid residues of the enzyme. The QSAR model built was stable, robust, and had a good predicting ability. Thus, predictions made by the model were reliably employed in further in silico studies. The compounds designed were more active than the template and showed better inhibition of the CDK2 enzyme compared to the standard drugs sorafenib and kenpaullone.

中文翻译：

---

对 HCT116 结直肠癌细胞系具有有效细胞毒活性的 CDK2 抑制剂的计算机设计和分子对接研究。

结直肠癌对于男女都很常见；发病率排名第三，死亡率排名第二，分别占全球男性和女性癌症病例的 10% 和 9.2%。尽管贝伐珠单抗、坎普托沙和西妥昔单抗等药物被用于治疗结直肠癌，但据报道，这些药物的疗效因患者而异。据报道，这些药物也有不同程度的副作用；因此，需要具有更好功效和更少副作用的新药物疗法。计算机药物设计方法为先导化合物的识别和优化提供了一种更快、更具成本效益的方法。因此，本研究的目的是通过计算机设计方法设计新型咪唑-5-酮。采用遗传函数算法方法建立QSAR模型来模拟化合物对HCT116结直肠癌细胞系的细胞毒性。建立的模型具有统计参数；R2 = 0.7397、R2adj = 0.6712、Q2cv = 0.5547 和 R2ext。= 0.7202，并揭示了化合物的细胞毒活性取决于分子描述符 nS、GATS5s、VR1_Dze、ETA_dBeP 和 L3i。这些分子描述符相关性较差（VIF < 4.0），并对构建的模型做出了独特的贡献。该模型用于通过基于配体的药物设计方法设计一组新颖的衍生物。与模板相比，化合物 e、h、j 和 l 显示出明显更好的细胞毒性 (IC50 < 5.0 μM)。通过分子对接研究研究了化合物与 CDK2 酶（PDB ID：6GUE）的相互作用。这些化合物是该酶的有效抑制剂，其结合亲和力范围为-10.8至-11.0 kcal/mol，并且主要与该酶的赖氨酸、天冬氨酸、亮氨酸和组氨酸氨基酸残基形成氢键相互作用。建立的QSAR模型稳定、鲁棒、具有良好的预测能力。因此，该模型做出的预测在进一步的计算机研究中得到了可靠的应用。与标准药物索拉非尼和肯保隆相比，设计的化合物比模板更具活性，并且对 CDK2 酶具有更好的抑制作用。