中文摘要：

蛋白激酶在信号转导、基因转录和蛋白翻译等生物过程起关键性作用,因而与大量人类疾病密切相关。所以,蛋白激酶的抑制剂筛选是抗肿瘤药物开发的热点,正在向基于全激酶组的高通量多靶点筛选模式发展。为了降低大规模实验筛选的成本,提高成功率,本文构建人类蛋白激酶组的多靶点分子对接系统,对抑制剂-激酶组的相互作用进行预测。我们首先利用同源模建方法,对人类激酶组约500个激酶变异体的催化域进行结构建模;接着以催化域结构模型为受体,用已知激酶抑制剂进行分子对接,对抑制剂与各激酶变异体的结合亲和力进行了定量计算。结果显示,本文所建立的多靶点分子对接系统可以准确预测抑制剂与激酶变异体的相互作用,结合自由能的计算值与实验值有很强的相关性。所以,该分子对接系统可用于多靶点激酶抑制剂的计算筛选,为激酶抑制剂开发与抗肿瘤药物设计提供理论依据。

英文摘要：

Protein kinases play critical roles in many biological processes, including signal transduction, gene transcription, and protein translation, and are therefore closely associated with various disease states. The screening of kinase inhibitors has become an important aspect of anti-tumor drug development, and has been refined to allow high-throughput, multi-target screening based on the entire human kinome. To reduce the experimental costs of large-scale inhibitor screening and to increase the success rate, our group has designed a multi-target molecular docking system capable of predicting kinase-inhibitor interactions. In this work we initially used homology modeling to construct three-dimensional （3D） models for approximately 500 catalytic domains of human kinase variants. We subsequently performed molecular docking to calculate the binding affinities of kinase-inhibitor pairs, employing the 3D models as receptors and kinase inhibitors as ligands. The results show that our multi-target docking system accurately predicts the interactions between known inhibitors and kinase variants, and that the calculated binding affinities are highly correlated with the experimental values. Thus, this molecular docking system could be used for computational screening of multi-target kinase inhibitors, thereby providing a theoretical basis for the development of kinase inhibitors and the design of anti-tumor drugs.