[Bioinformatics analysis of primary biliary cholangitis key genes and molecular mechanisms].

ABSTRACT

Objective: To extract the differentially expressed key genes of primary biliary cholangitis (PBC) using bioinformatics methods, so as to provide information for further study into the mechanism. Methods: The GSE119600 dataset was downloaded from the GEO database to obtain differentially expressed genes. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were performed for differentially expressed genes. Protein-protein interaction (PPI) network reconstruction, Cytoscape software visualization, and core gene screening were performed. The area under the receiver operating characteristic curve (ROC AUC) was used to assess the diagnostic effectiveness of genes and plot the pROC software package. The x-Cell software was used to calculate the enrichment score of 34 immune cells in each sample. Finally, four key genes (PSMA4, PSMA1, PSMB1, and PSMA3) were selected. Blood samples were analyzed using the qPCR method. Results: A total of 373 immune-related differentially expressed genes were identified. Eight genes (PSMC6, PSMB2, PSMB1, PSMA3, PSMA4, PSMA1, PSMD7, and PSMB5) were screened from the 178 nodes and 596 edges as hub genes of the PPI network, which were significantly related to amino acid metabolism, hematopoietic stem cell differentiation, cell cycle, and immune processes. PSMA4, PSMA1, PSMB1, and PSMA3 were defined as immunological biomarkers for PBC with an AUC value of the ROC curve > 0.7. Immunoinfiltrating cell analysis showed that the proportion of eosinophils was significantly higher in PBC patients compared to the control group, whereas the proportion of CD4+ memory T cells, plasma cells, Th2 cells, and cDC cells was significantly lower in PBC patients than the control group. Plasma cells were associated with all four immunological biomarkers. Seven PBC patients and seven healthy subjects were selected for peripheral blood qPCR validation, which demonstrates that PSMB1, PSMA3, PSMA1, and PSMA4 levels were significantly lower in PBC patients than healthy subjects, with a statistically significant difference. Conclusion: Bioinformatics screened eight key genes, of which four were key immunological markers and may serve as a basis for clinical diagnosis and mechanism exploration.