摘要
目前,治疗乳腺癌的候选药物是能够抗结ERα活性的化合物合成的,但由于化合物定量结构复杂、药代动力学性质(ADMET)不稳定,导致药物研发成本较高。本文通过相关性分析得出对ERα活性影响较高的20个分子描述符,并基于数据挖掘技术和机器学习算法,建立了相关化合物定量结构-ERα活性以及定量结构-ADMET性质的定量预测模型,对药物研发具有一定帮助。
At present, compounds with anti-junction ERα activity are drug candidates for the treatment of breast cancer, but due to the complex quantitative structure of the compound and the unstable pharmacokinetic properties (ADMET), the drug development cost is high. In this paper, 20 molecu-lar descriptors with high influence on ERα activity are obtained through correlation analysis, and based on data mining technology and machine learning algorithm, a quantitative prediction model of quantitative structure-ERα activity and quantitative structure-ADMET properties of related compounds is established, which is helpful for drug development.
出处
《建模与仿真》
2023年第1期130-139,共10页
Modeling and Simulation
