Multi-Level Latent Variable Models for Coheritability Analysis in Electronic Health Records

Area of Science:

• Genetics
• Biostatistics
• Computational Biology

Background:

• Electronic health records (EHRs) linked with familial data enable large-scale genetic architecture studies.
• Existing heritability methods struggle with complex family structures, diverse phenotypes, and scalability.
• Investigating shared genetic influences (coheritability) is crucial for understanding complex diseases.

Purpose of the Study:

• Develop a robust statistical framework for jointly estimating heritability and genetic correlation.
• Address limitations of existing methods in EHR-based family studies.
• Facilitate scalable coheritability analysis in high-dimensional data.

Main Methods:

• Utilize multi-level latent variable models to decompose phenotypic covariance.
• Incorporate within- and between-family variations for genetic and environmental components.
• Employ generalized estimating equations (GEE) for estimation algorithms.

Main Results:

• Simulation studies confirm estimator consistency and valid inference across various settings.
• Real-world EHR data analysis identified significant genetic correlations between mental health and endocrine/metabolic phenotypes.
• The framework demonstrates scalability for high-dimensional EHR data.

Conclusions:

• The proposed framework offers a rigorous and flexible approach for coheritability analysis in EHR data.
• Identified shared genetic etiology between mental health and metabolic conditions.
• Enables the discovery of shared genetic influences within complex disease networks.