Background. Tuberculosis (TB) is usually caused by Mycobacterium tuberculosis, which has the highest mortality rate among infectious diseases. This study is designed to identify the key genes affecting the diagnosis and treatment of TB. Methods. GSE54992, which included 39 peripheral blood mononuclear cell (PBMC) samples, was extracted from the Gene Expression Omnibus database. After the samples were classified into type and time groups by limma package, the differentially expressed genes (DEGs) were analyzed using the Analysis of Variance. Using pheatmap package, hierarchical cluster analysis was performed for the DEGs. Then, the key modules correlated with TB were selected using the WGCNA package. Finally, functional and pathway enrichment analyses were carried out using clusterProfiler package. Results. The DEGs in subclusters 3, 6, 7, and 8 were chosen for further analyses. Based on WGCNA analysis, blue and green modules in type group and pink module in time group were selected as key modules. From the key modules, 9 (including BAX and ARPC1B) hub genes in type group and 6 (including DHX36) hub genes in time group were screened. Through pathway enrichment analysis, the TNF signaling pathway was enriched for the green module. Conclusion. DHX36, BAX, and ARPC1B might be key genes acting in the mechanisms of TB. Besides, the TNF signaling pathway might also be critical for the diagnosis and therapy of the disease.