Identification of Potential Feature Genes in CRSwNP Using Bioinformatics Analysis and Machine Learning Strategies

Abstract

Huikang Wang,1– 3,* Xinjun Xu,1– 3,* Haoran Lu,1– 3 Yang Zheng,1– 3 Liting Shao,1– 3 Zhaoyang Lu,2– 4 Yu Zhang,2,3 Xicheng Song2,3 1Department of Otorhinolaryngology, Head and Neck Surgery, Yantai Yuhuangding Hospital, QingdaoUniversity, Yantai, People’s Republic of China; 2Shandong Provincial Clinical Research Center for Otorhinolaryngologic Diseases, Yantai Yuhuangding Hospital, Yantai, People’s Republic of China; 3Yantai Key Laboratory of Otorhinolaryngologic Diseases, Yantai Yuhuangding Hospital, Yantai, People’s Republic of China; 4Second Clinical Medicine College, Binzhou Medical University, Yantai, Shandong, 264003, People’s Republic of China*These authors contributed equally to this workCorrespondence: Xicheng Song; Yu Zhang, Department of Otolaryngology, Head and Neck Surgery. Yantai Yuhuangding Hospital, Qingdao University, Yantai, 264000, People’s Republic of China, Tel [phone], Fax [phone], Email [email protected]; [email protected]: The pathogenesis of CRSwNP is complex and not yet fully explored, so we aimed to identify the pivotal hub genes and associated pathways of CRSwNP, to facilitate the detection of novel diagnostic or therapeutic targets.Methods: Utilizing two CRSwNP sequencing datasets from GEO, differential expression gene analysis, WGCNA, and three machine learning methods (LASSO, RF and SVM-RFE) were applied to screen for hub genes. A diagnostic model was then formulated utilizing hub genes, and the AUC was generated to evaluate the performance of the prognostic model and candidate genes. Hub genes were validated through the validation set and qPCR performed on normal mice and CRSwNP mouse model. Lastly, the ssGSEA algorithm was employed to assess the differences in immune infiltration levels.Results: A total of 239 DEGs were identified, with 170 upregulated and 69 downregulated in CRSwNP. Enrichment analysis revealed that these DEGs were primarily enriched in pathways related to nucleocytoplasmic transport and HIF-1 signaling pathway. Data yielded by WGCNA analysis contained 183 DEGs. The application of three machine learning algorithms identified 11 hub genes. Following concurrent validation analysis with the validation set and qPCR performed after establishing the mouse model confirmed the overexpression of BTBD10, ERAP1, GIPC1, and PEX6 in CRSwNP. The examination of immune cell infiltration suggested that the infiltration rate of type 2 T helper cell and memory B cell experienced a decline in the CRSwNP group. Conversely, the infiltration rates of Immature dendritic cell and Effector memory CD8 T cell were positive correlation.Conclusion: This study successfully identified and validated BTBD10, ERAP1, GIPC1, and PEX6 as potential novel diagnostic or therapeutic targets for CRSwNP, which offers a fresh perspective and a theoretical foundation for the diagnostic prediction and therapeutic approach to CRSwNP.Keywords: chronic rhinosinusitis with nasal polyposis, key genes, machine learning, immune cell infiltration

Keywords

• chronic rhinosinusitis with nasal polyposis
• key genes
• machine learning
• immune cell infiltration