Exploring of the shared gene signatures and molecular mechanism in COVID-19 and tuberculosis

Abstract

Abstract Background After the Coronavirus Disease 2019 (COVID-19) pandemic, tuberculosis (TB) incidence has demonstrated a noticeable upswing, with the causative linkage and mechanistic crosstalk between these conditions remaining uncharted. This study endeavours to decipher the communal genetic elements and underlying molecular interplayunderlying COVID-19 and TB. Methods The Gene Expression Omnibus (GEO) served as the repository for sourcing RNA sequence datasets pertinent to COVID-19 and TB. Leveraging the R software, the Weighted Gene Co-expression Network Analysis (WGCNA) and limma package facilitated the uncovering of a co-expression network intertwined with both COVID-19 and TB. Shared genes underwent enrichment analysis via ClueGO, while hub genes within the COVID-19 and TB context were identified through MCODE based on Cytoscape software. An array of machine learning algorithms – Random Forests (RF), Least Absolute Shrinkage and Selection Operator (LASSO) Logistic Regression, and Support Vector Machine-Recursive Feature Elimination (SVM-RFE) – guided the further isolation of key genes. We also constructed the nomograms, and assessed the predictive prowess by evaluating the Area under the curve (AUC), calibration curves, decision curve analysis (DCA) and clinical impact curves. The immune microenvironment (TIME) in TB was analyzed using CIBERSORT, allowing for the assessment of correlation between key genes and immune cells. Results WGCNA analyses and gene expression differences analysis based on the Limma divulged a set of 281 shared differential genes between TB and COVID-19. Enrichment analysis elucidated their association with a variety of biological functions and signaling pathways, such as response to interferon-γ, NOD-like receptor signaling pathway, and influenza A. Machine learning facilitated the identification of GBP5 and IFITM3 genes, which were subsequently fashioned into nomograms, exhibiting solid clinical relevance (AUC = 0.9854, Mean Absolute Error = 0.009). CIBERSORT analysis uncovered substantial shifts in multiple immune cells in TB, notably Macrophages.M1, Dendritic.cells.activated, and Neutrophils cells, which revealed strong correlation with the expression of GBP5 and IFITM3 in TB. Conclusion In this study, we obtained the shared gene between COVID-19 and TB and preliminarily revealed its function. In addition, GBP5 and IFITM3 could serve as key genes among the shared genes and were associated with a variety of immune cells.