Variational Autoencoder-based Model Improves Polygenic Prediction in Blood Cell Traits.

Area of Science:

• Genomics
• Computational Biology
• Bioinformatics

Background:

• Large-scale genomic studies enable genetic predisposition assessment.
• Polygenic Risk Scores (PRS) aggregate genomic information for personalized risk prediction.
• Conventional PRS methods using linear models struggle with high-dimensional genomic data and complex interactions.

Purpose of the Study:

• To improve the predictive power of Polygenic Risk Scores (PRS) using advanced deep learning techniques.
• To develop a novel deep learning model for enhanced genetic risk prediction.
• To capture complex genetic patterns and interaction effects for more accurate trait prediction.

Main Methods:

• Application of a Variational AutoEncoder-based model for PRS construction (VAE-PRS).
• Utilizing deep learning techniques to analyze high-dimensional genomic data.
• Employing SHapley Additive exPlanations (SHAP) for model interpretability.

Main Results:

• VAE-PRS outperformed state-of-the-art methods in 14 out of 16 blood cell traits using biobank-level data.
• The model demonstrated computational efficiency and robustness across different variant sets.
• VAE-PRS effectively captured interaction effects in high-dimensional genomic data.
• SHAP analysis provided interpretability by assessing individual marker contributions.

Conclusions:

• VAE-PRS offers a novel and powerful deep learning approach for genetic risk prediction.
• The model enhances the predictive accuracy of Polygenic Risk Scores.
• VAE-PRS facilitates personalized medicine and genetic research through improved interpretability and performance.