Age-related hearing loss (ARHL) affects 1 in 3 adults over 65, reducing quality of life and increasing dementia risk. Current diagnostics rely on late-stage symptoms, and no disease-modifying treatments exist. While genetics and environment contribute to ARHL, the molecular mechanisms remain poorly understood. Recent advances in multi-omics technologies (genomics, proteomics, metabolomics) offer unprecedented opportunities to identify early predictors and therapeutic targets. The UK Biobank’s large-scale dataset-combining auditory tests, genetic data, and blood biomarkers-provides a unique platform to address this gap.
The objective of this study is to develop a machine learning model that predicts the risk of ARHL using clinical, genetic, and biochemical data. Furthermore, we will integrate multi-omics data, including genomics, proteomics, and metabolomics, to uncover dysregulated pathways and potential target proteins that can facilitate their translation into future clinical applications.
We will leverage the resources of the UK Biobank by extracting pure-tone audiometry results to define the severity of ARHL and use clinical data, genomics, proteomics/serum biomarkers, and metabolomics data as predictors. We will employ machine learning modeling methods to train predictive models and validate their performance. Additionally, we will use multi-omics integration approaches to link genomic variants, protein levels, and metabolic shifts to ARHL subgroups through network analysis and prioritize druggable targets.
The validated ARHL risk assessment tool developed in this study is expected to enable early interventions, such as providing hearing aids before irreversible damage occurs. Moreover, the identified target proteins may accelerate the repurposing of existing drugs, like anti-inflammatory agents, and expedite the development of new therapies.
