Identification of Protein Markers Predictive of Drug-Specific Survival Outcome in Cancers
Identification of Protein Markers Predictive of Drug-Specific Survival Outcome in Cancers
复制标题
DOI：
10.1007/978-3-030-91415-8_6
复制
发表时间：
2021
影响因子：
9.2
通讯作者：
Shuting Lin;Jie-Chao Zhou;Yiqiong Xiao;B. Neary;Yong Teng;Peng-Chao Qiu
中科院分区：
文献类型：
--
作者：
Shuting Lin;Jie-Chao Zhou;Yiqiong Xiao;B. Neary;Yong Teng;Peng-Chao Qiu
预测药物特异性反应的生物标志物的新发现不仅在揭示癌症的药物机制方面发挥了关键作用，而且对个性化药物也至关重要。在这项研究中，我们通过整合来自癌症基因组图谱(TCGA)的7076名患者的蛋白质表达数据、药物治疗数据和生存结果来识别药物特异性生物标记物。我们首先定义了抗癌药物组，其中每个抗癌药物组包含患有相同癌症并接受相同药物治疗的患者。对于每种蛋白，我们根据蛋白的高表达或低表达对每个癌症药物组的患者进行分层，并应用对数等级检验来检验分层患者是否表现出显著的生存差异。我们检查了98组癌症药物中的336种蛋白质，确定了65种蛋白质-癌症-药物组合，涉及55种独特的蛋白质，其中蛋白质表达水平可以预测药物特定的生存结果。一些已鉴定的蛋白质得到了已发表的文献的支持。利用TCGA的基因表达数据，我们发现11%的药物特异性蛋白的mRNA表达也与药物特异性生存显著相关，并且这些药物特异性蛋白及其对应的基因大多具有很强的相关性。
Novel discoveries of biomarkers predictive of drug-specific responses not only play a pivotal role in revealing the drug mechanisms in cancers, but are also critical to personalized medicine. In this study, we identified drug-specific biomarkers by integrating protein expression data, drug treatment data and survival outcome of 7076 patients from The Cancer Genome Atlas (TCGA). We first defined cancer-drug groups, where each cancer-drug group contains patients with the same cancer and treated with the same drug. For each protein, we stratified the patients in each cancer-drug group by high or low expression of the protein, and applied log-rank test to examine whether the stratified patients show significant survival difference. We examined 336 proteins in 98 cancer-drug groups and identified 65 protein-cancer-drug combinations involving 55 unique proteins, where the protein expression levels are predictive of drug-specific survival outcomes. Some of the identified proteins were supported by published literature. Using the gene expression data from TCGA, we found the mRNA expression of11% of the drug-specific proteins also showed significant correlation with drug-specific survival, and most of these drug-specific proteins and their corresponding genes are strongly correlated.