Integrated bioinformatics and machine learning deciphering IL12RB2 and FYN as key immune biomarkers in brucellosis

Brucellosis remains a major zoonotic threat worldwide, and early diagnosis is challenging due to nonspecific symptoms and complex immune evasion strategies. Immune-related biomarkers may provide novel diagnostic and therapeutic clues, yet they remain poorly characterized in brucellosis.

This study aimed to identify and validate immune-related biomarkers linked to brucellosis, determine their diagnostic utility, and examine their potential as candidates for further therapeutic investigation, alongside analyses of immune pathways and immune-cell infiltration patterns.

Transcriptome data from patients with brucellosis and healthy controls (GSE69597) were integrated with 1793 immune-related genes from ImmPort. Differential expression analysis and weighted gene co-expression network analysis (WGCNA) identified candidate genes, which were further screened using three machine-learning algorithms (LASSO, random forest, and SVM-RFE). Immune pathway enrichment, immune-cell infiltration (CIBERSORT), and molecular docking with first-line drugs (doxycycline and rifampicin) were performed. qRT-PCR was used to validate hub gene expression in PBMCs from clinical cohorts.

RFX5, FYN, IL12RB2, and LGR6 demonstrated high diagnostic value in ROC analysis. Enrichment analyses indicated that brucellosis-related IRGs are primarily involved in immune recognition, activation of immune responses, and regulation of inflammatory processes, and were significantly correlated with levels of various immune cell infiltrates. Exploratory molecular docking suggested potential binding propensities between standard anti-Brucella antibiotics (doxycycline and rifampicin) and the candidate host proteins. Clinical sample validation further confirmed significant expression changes of IL12RB2 and FYN in patients with brucellosis.

IL12RB2 and FYN serve as robust immune biomarkers for human brucellosis and may represent therapeutic candidates linked to Th1 signaling and T-cell activation. This integrative computational-experimental framework provides a foundation for precision diagnosis and immunomodulatory strategies in brucellosis.