Biomarker host genetic signatures are considered key tools for improved early diagnosis of tuberculosis (TB) disease (development). The analysis of gene expression changes based on a limited number of genes or single study designs, however, may not be sufficient for the identification of universal diagnostic biomarker profiles. Here we propose that biological pathway and process based analyses from multiple data sets may be more relevant for identification of key pathways in TB pathogenesis, and may reveal novel candidate diagnostic TB biomarkers. A number of independent genome-wide gene expression studies have recently been performed to study expression of biomarkers for TB disease. We have integrated the results from these independent studies and performed pathway- as well as biological process-based analysis on the total data set. Interestingly, IFNα/β signalling is not the single dominant pathway in the analysis of the total dataset, but combined, functional, analysis of biomarkers suggests a strong dominant role for myeloid cell involvement in inflammation.