Integrative multi-omics analysis identifies SELL and PRF1 as key immune biomarkers and therapeutic targets in tuberculosis.

Tuberculosis (TB), caused by Mycobacterium tuberculosis (M.tb), remains a major global health burden, with host immune responses critically influencing disease progression and treatment outcomes. In this study, we performed an integrative multi-cohort analysis combining transcriptomics, SNP-based structural evaluation, and molecular docking to identify key immune regulators and potential therapeutic targets. Transcriptomic datasets from macrophage and patient-derived samples were analyzed to identify differentially expressed genes, followed by functional enrichment and network-based analyses to identify key immune regulators. Non-synonymous SNPs (nsSNPs) were evaluated using multiple predictive tools (SIFT, PolyPhen-2, CADD, REVEL, MetaLR, Mutation Assessor), while structural impacts were assessed using DUET and Project HOPE. A total of 162 DEGs were identified, significantly enriched in immune-related pathways including cytokine signaling, receptor activity, and host defense responses. Five hub genes SELL, CD19, CD27, PRF1, and KLRK1 were prioritized, with SELL and PRF1 demonstrating the highest regulatory importance. Structural and functional analyses identified six deleterious nsSNPs in SELL (G169E, W309R, W247R) and PRF1 (V419G, V482G, P459S), all predicted to destabilize protein structure and impair immune function. Molecular docking and MM/GBSA analyses revealed favorable and energetically stable interactions between selected immunomodulatory compounds and both wild-type and mutant protein structures. Key ligand-residue interactions supported the potential of these compounds to modulate SELL- and PRF1-associated immune pathways. Notably, mutant variants showed enhanced ligand binding, with BIMOSIAMOSE displaying the highest affinity in SELL W309R (- 6.92 kcal/mol) and PRF1 mutants (- 5.35 kcal/mol), suggesting mutation-induced alterations in binding pocket conformation. Unlike previous single-cohort or single-method studies in tuberculosis, this work employs a multi-cohort, systems-level integrative framework to identify robust host immune signatures. By combining transcriptomic meta-analysis across diverse infection models with protein-protein interaction network prioritization, regulatory network inference, genetic variant impact assessment, and structure-based computational validation, we identified conserved immune-associated genes, particularly SELL and PRF1, as central regulatory nodes in tuberculosis pathogenesis. The integration of structural modeling, molecular docking, and molecular dynamics simulations further strengthens the mechanistic plausibility of these candidates. This multi-layered strategy provides a comprehensive framework for prioritizing biologically consistent biomarkers and therapeutic targets in tuberculosis, offering a systems-level perspective beyond traditional single-omics approaches.