广东药科大学学报2025，Vol.41Issue(3)：52-56,5.DOI:10.16809/j.cnki.2096-3653.2024121001
基于基因表达的吉非替尼与奥西替尼的药物敏感性预测模型构建
Construction of a gene expression-based drug sensitivity prediction model for Gefitinib and Osimertinib
摘要
Abstract
Objective To identify genes associated with sensitivity to Gefitinib and Osimertinib based on machine learning and develop predictive models for drug sensitivity.Methods Gene expression profiles were obtained from the Cancer Genome Atlas(TCGA)and the Gene Expression Omnibus(GEO),and corresponding expression profiles and IC50 values were retrieved from the GDSC database to facilitate the identification of genes related to Gefitinib and Osimertinib sensitivity.Prediction models for the sensitivity to both drugs were constructed based on Lasso,logistic regression,and random forest algorithms,and key genes correlated with drug sensitivity were identified.Results A total of 28 genes associated with Gefitinib drug sensitivity and 84 genes linked to Osimertinib drug sensitivity were obtained.The predictive efficacy of gefitinib drug sensitivity prediction model was RF:AUC=0.825,LR:AUC=0.812,and the prediction effect of the osimertinib drug sensitivity prediction model was RF:AUC=0.850,LR:AUC=0.817.Conclusion Sensitivity models for Gefitinib and Osimertinib have been constructed by utilizing machine learning algorithms.Hub genes such as TMEM158 and RAB3B have then been identified to predict drug sensitivity.关键词
吉非替尼/奥西替尼/机器学习/药物敏感性Key words
Gefitinib/Osimertinib/machine learning/drug sensitivity分类
药学引用本文复制引用
王祎允,马克威..基于基因表达的吉非替尼与奥西替尼的药物敏感性预测模型构建[J].广东药科大学学报,2025,41(3):52-56,5.基金项目
北京医学奖励基金会课题研究项目(YXJL-2024-0129-0004) （YXJL-2024-0129-0004）
吉林大学第一医院临床研究专项项目(3R218R613428) （3R218R613428）
