Construction and validation of an immune gene-based model for diagnosis and risk prediction of severe asthma.

Abstract

Objective: Severe asthma (SA) is a serious disease with limited treatment options, which is closely linked to immune dysfunction. Therefore, immune-associated biomarkers may diagnose SA and offer therapeutic targets for SA.

Methods: The gene expression profiles of SA patients and matched controls were from the National Center for Biotechnology Information database. Immune genes were downloaded from the ImmPort database. After screening for differentially expressed genes (DEGs) between SA patients and controls, and identifying gene modules highly associated with SA, immune-related DEGs were obtained. Then, protein-protein interaction analysis, Cytoscape software and receiver operating characteristic (ROC) curves were used to identify hub genes. Next, the relationship between hub genes and immune cells was explored, and single-sample gene set enrichment analysis (ssGSEA) was applied to conduct pathway enrichment analyses. Finally, the Least Absolute Shrinkage and Selection Operator (LASSO) combined with ROC analysis were used to confirm the diagnostic value of the hub genes.

Results: Forty immune-related DEGs were obtained, and RNASE3, CAMP and LTF were determined as hub genes. The hub genes were closely associated with immune cells, and ssGSEA showed that lysosome was associated with high expressions of the hub genes, while primary immunodeficiency was related to low expressions of the hub genes. LASSO combined with ROC analysis confirmed the immune gene-based model (RNASE3, CAMP, LTF, and CD79A) could distinguish SA patients from healthy individuals with high sensitivity.

Conclusions: RNASE3, CAMP, LTF, and CD79A could act as diagnostic markers for SA, providing a theoretical basis for developing diagnostic targets for SA.