TIGIT and SNRPA1 as novel diagnostic and predictive biomarkers in obstructive ventilatory dysfunction combined with pulmonary nontuberculous mycobacterial infection patients.

Background: Patients with obstructive ventilatory dysfunction are prone to NTM (nontuberculous mycobacterial) colonization and infection. Our investigation employs an integrative bioinformatics approach to elucidate critical molecular signatures linked to obstructive ventilatory dysfunction combined with NTM infection, and constructing a clinical diagnostic model using core differentially expressed genes. Methods: The GSE97298 dataset from the GEO database was analyzed via GEO2R to identify differentially expressed genes (DEGs). Enrichment analysis of the DEGs was conducted using Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses and Gene Set Enrichment Analysis (GSEA). DEGs were further conducted immune infiltration analysis and screened for core genes through protein-protein interaction (PPI) network analysis and machine learning (Lasso regression and Random Forest). The DSigDB database was employed to explore the potential targeted drugs of characteristic genes. Diagnostic potential and predictive model of candidate biomarkers was assessed through five machine learning (ML) techniques, namely XGBoost, LightGBM, Random Forest (RF), Adaptive Boosting (Adaboost), and Support Vector Machine (SVM) modeling. Shapley Additive Interpretation (SHAP) analysis facilitated a visual interpretation for individual patient. Results: A total of 69 DEGs, which are widely involved in important biological processes such as cell cycle, tubulin binding, and RNA processing, were identified. Immune cell analysis indicated that B cells, T follicular helper cells and T Cells were positively correlated with NTM, Mast cells, Macrophages cells and NK CD56dim cells showed a negative correlation. Through PPI network analysis and machine learning, TIGIT and SNRPA1 were selected as the core gene and significant predictors for subsequent analysis. The expression of TIGIT and SNRPA1 proteins in NTM patients' blood samples were also down-regulated compared with control. Using the DSigDB database, we predicted seven drugs that exhibit significant binding activity with core genes. Importantly, SNRPA1/TIGIT was the optimal combination for predicting obstructive ventilatory dysfunction combined with NTM infection. The RF model surpassing the performance of other models. SHAP analysis provided independent explanations, reaffirming the critical factors associated with the risk of obstructive ventilatory dysfunction combined with NTM infection. Conclusion: Our study demonstrated that TIGIT and SNRPA1 were down-regulated and were strongly associated with NTM infection. In addition, we successfully established a precise predictive model for risk of obstructive ventilatory dysfunction combined with NTM infection using machine learning techniques, offering valuable support to clinicians in making informed clinical decisions.