Pharmacogenomic Cluster Analysis of Lung Cancer Cell Lines Provides Insights into Preclinical Model Selection in NSCLC

• Published in: Interdisciplinary sciences : computational life sciences Vol. 14; no. 3; pp. 712 - 721
• Main Authors: Shen, Yueyue, Xiang, Ying, Huang, Xiaolong, Zhang, Youhua, Yue, Zhenyu
• Format: Journal Article
• Language: English
• Published: Singapore Springer Nature Singapore 01.09.2022; Springer Nature B.V
• Subjects: Antitumor agents; Biomarkers; Biomedical and Life Sciences; Cell differentiation; Cellular stress response; Chemotherapy; Cluster analysis; Clustering; Computational Biology/Bioinformatics; Computational Science and Engineering; Computer Appl. in Life Sciences; Consistency; Copy number; Drug screening; Genomes; Health Sciences; Life Sciences; Lung cancer; Lung carcinoma; Mathematical and Computational Physics; Medicine; Mutation; Non-small cell lung carcinoma; Original Research Article; Oxidative stress; Pharmacogenomics; Small cell lung carcinoma; Squamous cell carcinoma; Statistics for Life Sciences; Theoretical; Theoretical and Computational Chemistry; Tumor cell lines; Tumors; Cluster analysis; Preclinical model selection; Cell line; Non-small cell lung cancer; Pharmacogenomics
• ISSN: 1913-2751; 1867-1462
• DOI: 10.1007/s12539-022-00517-z

Human lung cell lines are utilized widely for investigating tumor biology, experimental therapy, anticancer drug screening and biomarkers identification. However, the consistency of drug responses of these established cell lines and non-small cell lung cancer (NSCLC) is uncertain. In this study, we assessed the drug response consistency between lung cell lines and NSCLC tumors in The Cancer Genome Atlas by hierarchical clustering using copy number variations in driver genes, and profiled the molecular patterns and correlations in cell lines. We found that some frequently used cell lines of NSCLC subtypes were not clustered with their matched subtypes of tumor. Mutation profiles in the oxidative stress response and squamous differentiation pathway in lung cell lines were in concordance with lung squamous cell carcinoma. Furthermore, lung cell lines and tumors in the same sub-cluster had very similar responses to certain drugs but some were inconsistent, suggesting that clustering through copy number variation data could capture part of the suitability of lung cell lines. The analysis of these results could aid investigators in evaluating drug response models and eventually enabling personalized treatment recommendations for individual patients with NSCLC. Graphical abstract