背景：顺铂耐药在肺癌中是一个常见的临床问题。然而，其潜在机制尚未完全阐明，突出了寻找生物靶点的重要性。 方法：通过下载的公共数据（GSE21656、GSE108214和TCGA）和特定的R软件包完成生物信息学分析。通过CCK8试验、细胞集落形成试验和EdU试验完成细胞增殖能力的评估。通过Transwell和刮伤愈合试验完成细胞侵袭和迁移能力的评估。此外，通过计算IC50评估细胞对顺铂的敏感性。 结果：在这里，我们发现PCDH7可能通过公共数据库分析（GSE21656和GSE108214）参与肺癌中的顺铂耐药。然后，我们进行了一系列的体外实验，验证了PCDH7在非小细胞肺癌（NSCLC）中的促癌作用。此外，IC50检测结果显示PCDH7可能与NSCLC的顺铂耐药有关。接下来，我们研究了PCDH7的单细胞模式、生物学功能和免疫分析。重要的是，我们注意到PCDH7可能调控上皮-间质转化活性以及CD8+T细胞和活化NK细胞的局部浸润。此外，我们还注意到高PCDH7表达的患者可能更敏感于硼替佐米、多西他赛和吉西他滨，并对免疫治疗产生抵抗。最后，我们通过机器学习算法构建了基于三个PCDH7相关基因（GPX8、BCAR3和TNS4）的预后模型，对NSCLC患者的生存具有很好的预测能力。 结论：我们的研究改进了NSCLC中顺铂耐药的调控框架，可以为后续相关研究提供指导，特别是关于PCDH7。 版权所有 © 2023 Li、Xu、Guo、Du和Chen。

Background: Cisplatin resistance is a common clinical problem in lung cancer. However, the underlying mechanisms have not yet been fully elucidated, highlighting the importance of searching for biological targets. Methods: Bioinformatics analysis is completed through downloaded public data (GSE21656, GSE108214, and TCGA) and specific R packages. The evaluation of cell proliferation ability is completed through CCK8 assay, colony formation, and EdU assay. The evaluation of cell invasion and migration ability is completed through transwell and wound-healing assays. In addition, we evaluated cell cisplatin sensitivity by calculating IC50. Results: Here, we found that PCDH7 may be involved in cisplatin resistance in lung cancer through public database analysis (GSE21656 and GSE108214). Then, a series of in vitro experiments was performed, which verified the cancer-promoting role of PCDH7 in NSCLC. Moreover, the results of IC50 detection showed that PCDH7 might be associated with cisplatin resistance of NSCLC. Next, we investigated the single-cell pattern, biological function, and immune analysis of PCDH7. Importantly, we noticed PCDH7 may regulate epithelial-mesenchymal transition activity, and the local infiltration of CD8+ T and activated NK cells. Furthermore, we noticed that patients with high PCDH7 expression might be more sensitive to bortezomib, docetaxel, and gemcitabine, and resistant to immunotherapy. Finally, a prognosis model based on three PCDH7-derived genes (GPX8, BCAR3, and TNS4) was constructed through a machine learning algorithm, which has good prediction ability on NSCLC patients' survival. Conclusion: Our research has improved the regulatory framework for cisplatin resistance in NSCLC and can provide direction for subsequent related research, especially regarding PCDH7.Copyright © 2023 Li, Xu, Guo, Du and Chen.