Machine Learning and Mendelian Randomization Identify Allergic Rhinitis as Nasopharyngeal Carcinoma Risk Factor With Validated Potential Candidate Biomarkers.

Background: Despite extensive research, nasopharyngeal carcinoma (NPC) remains a complex malignancy with poorly understood cellular dynamics and risk factors. The relationship between allergic rhinitis and NPC development has been controversial. This study combined single-cell transcriptomics with Mendelian randomization to comprehensively map cellular heterogeneity and establish potential causal links between allergic rhinitis and NPC pathogenesis. Methods: Researchers performed single-cell RNA sequencing on NPC and adjacent normal tissue samples. Simultaneously, a two-sample Mendelian randomization approach was employed using allergic rhinitis-associated genetic variants as instrumental variables to investigate causality between allergic rhinitis and NPC risk. Integration of these genetic instruments with transcriptomic profiles enabled the identification of genetically influenced molecular pathways. Results: Comprehensive analysis revealed intricate cellular landscapes comprising epithelial, immune, endothelial, and stromal cell populations, each demonstrating unique transcriptional signatures. Mendelian randomization analysis provided evidence for a causal relationship between allergic rhinitis and NPC development (OR = 1.42, 95% CI: 1.18-1.76, Conclusion: By leveraging both single-cell transcriptomics and Mendelian randomization, this study provides unprecedented insights into NPC's cellular complexity and establishes causal pathways linking allergic rhinitis to NPC development. These findings identify genetically validated molecular mechanisms that could represent promising therapeutic targets for this challenging malignancy.