Classifying Hearing-Loss-Causing Variants using Protein Structure and Artificial Intelligence (AI)

The development of treatments for hereditary diseases hinges on accurately identifying the specific genetic variants responsible for the condition. Current pathogenicity prediction tools often overlook the structural changes caused by mutations or fail to focus on disease-specific genes, limiting their predictive reliability and the depth of insight they can offer to scientists and clinicians. Our research focuses on the prediction and analysis of the pathogenicity of variants in hearing loss related genes, at the protein level. We leverage state-of-the-art AI methodologies and innovative features derived from proteins’ 3D structure, to assess the impact of genetic variants. This research is part of the big-data precision medicine project, enabling the discovery of novel disease-causing variants and driving the development of personalized therapies.