Blood transcriptomic profiling identifies hub genes with diagnostic and clinical relevance in non-tuberculous mycobacterial pulmonary disease

The global incidence of non-tuberculous mycobacterial pulmonary disease (NTM-PD) has surged in recent decades, establishing it as a formidable challenge in respiratory medicine. Despite its escalating clinical burden, current diagnostic approaches—primarily reliant on microbiological culture and species identification—remain constrained by prolonged turnaround times (3–6 weeks) and limited sensitivity. To address this critical gap, we aimed to identify and validate peripheral blood mRNA signatures as non-invasive diagnostic biomarkers for NTM-PD. We recruited 28 participants in the discovery phase, comprising 8 patients with Mycobacterium abscessus (MAB)-PD, 8 patients with Mycobacterium avium complex (MAC)-PD, and 12 age/sex-matched healthy controls (HC). Peripheral blood samples from all participants underwent bulk RNA sequencing and single-cell transcriptome sequencing (scRNA-seq). Hub genes prioritized through integrated bioinformatics analysis were subsequently validated in an expanded validation cohort of 50 NTM- PD patients (23 MAC-PD, 27 MAB-PD) and 35 HC using quantitative reverse transcription PCR (RT- qPCR). Finally, we assessed the correlation between hub gene expression levels and clinical parameters. Transcriptomic profiling revealed significant upregulation of eight core genes (IFIT1, IFI44L, GBP5, FOLR3, LY6E, IFITM1, S100A8, S100A12) in NTM-PD patients compared to healthy controls ( P < 0.05). Among these, four genes exhibited the most pronounced differential expression: IFIT1 ( P = 0.005), IFITM1 ( P = 0.001), S100A12 ( P < 0.001) and GBP5 ( P = 0.001). Importantly, these four hub genes demonstrated consistent discriminatory power across both Mycobacterium avium complex (MAC)-PD and M. abscessus (MAB)-PD subtypes. Single-cell transcriptomic analysis further confirmed that these hub genes were predominantly upregulated in monocyte/macrophage and neutrophil subpopulations, validating the bulk-level findings at cellular resolution. Receiver operating characteristic (ROC) curve analysis further identified the triple-gene panel (IFITM1 + S100A12+GBP5) as the optimal diagnostic signature, achieving an area under the curve (AUC) of 0.85 with 84% sensitivity and 78% specificity, significantly outperforming other candidate combinations. This study presents a comprehensive analysis of peripheral blood transcriptomic signatures in patients with non-tuberculous mycobacterial pulmonary disease (NTM-PD). By integrating bulk RNA-seq discovery with scRNA-seq cellular validation, we identified a three-gene diagnostic panel (IFITM1, S100A12 and GBP5) that robustly discriminates NTM-PD from healthy controls (AUC = 0.85). Single-cell analysis confirmed that these hub genes were predominantly upregulated in monocyte/macrophage and neutrophil subpopulations, validating the bulk-level findings at cellular resolution. Importantly, this biomarker panel offers a rapid, minimally invasive diagnostic tool with potential to transform clinical practice in resource-limited settings where conventional microbiological methods remain challenging.