IFNG and IFNGR1 polymorphisms are associated with tuberculosis: a case-control study

Previous studies suggested that single-nucleotide polymorphisms (SNPs) of interferon gamma (IFNG) and its receptor IFNGR1 may be involved in the pathogenesis of tuberculosis (TB). We aimed to examine the association of IFNG gene polymorphisms with TB in the Tibetan population and use the machine learning method to establish a clinical prediction model of TB.

A total of 613 TB patients and 603 healthy controls were selected for the study. Associations between SNPs and TB were analyzed using logistic regression, adjusted for sex and age. Clinical data and SNPs were integrated to construct a TB prediction model using random forest (RF) machine learning.

For IFNG, rs1861494 CT was a protective factor against TB compared with TT genotype (p = 0.010). The rs1861494 C allele was a protective factor for TB (p = 0.010). For IFNGR1, the rs3799488 C allele reduced the risk of TB by 30% (p < 0.001). rs9376267 CT (p = 0.005) and TT (p = 0.001) genotypes were protective factors for TB. Compared with the rs1327475 GG genotype, the frequency of the GA genotype in the case group significantly differed from the controls (p = 0.013). rs2234711 GA (p < 0.001), AA (p < 0.001) genotype and A (p < 0.001) alleles were also associated with TB. Finally, five markers are identified using the RF model. The area under the curve (AUC) reaches 0.6 in the training set and 0.59 in the test set.

Our study found that IFNG and IFNGR1 gene polymorphisms were associated with TB in a Tibetan population. The results also demonstrate the potential of clinical-SNPs as diagnostic tools for TB.