Multi-omics analyses prioritize potential biomarkers and drug targets for obstructive sleep apnea

Obstructive sleep apnea (OSA) remains a prevalent sleep disorder with incompletely characterized molecular mechanisms and evolving pharmacological options. We aimed to identify molecular biomarkers associated with OSA and prioritize candidates for further evaluation.

OSA genome-wide association study summary statistics were analyzed with blood- and brain-derived expression and protein quantitative trait loci (QTL) data using transcriptome-wide and proteome-wide association studies. Summary data-based Mendelian randomization (MR), HEIDI testing and Bayesian colocalization were then used to evaluate the identified genes and proteins. Candidate biomarkers were assessed using alternative QTL resources and OSA GWAS data. Multivariable MR assessed eligible associations after accounting for body mass index (BMI). Functional enrichment, methylation-related analyses, and pharmacological analyses provided downstream characterization.

28 unique genes and proteins were prioritized across blood/plasma and brain. 18 candidate biomarkers were supported using alternative QTL resources and the independent OSA GWAS dataset. Among assessable biomarkers, MAP7D1, MCM6, L3MBTL2, and LCT remained associated with OSA after accounting for BMI, whereas PCSK1 and most other associations were attenuated. Enrichment analyses implicated glucose metabolism, autophagy-lysosome processes and apoptosis-related pathways. Mediation analyses identified 43 significant indirect effects of 41 CpG sites on OSA through gene expression or protein abundance. Drug-database screening identified candidate drugs for several biomarkers, while molecular docking supported target-compound interactions involving BID and GPD2.

This study prioritized 28 molecular biomarkers associated with OSA and further characterized their biological and regulatory context, while selected candidates warrant further experimental evaluation.