Heritability Estimation using Regression Models for Correlation

Area of Science:

• Genetics
• Biostatistics

Background:

• Heritability estimation quantifies the polygenic effect on population traits.
• Accurate heritability interpretation is vital for genetic studies aiming to identify trait-associated genes.

Purpose of the Study:

• To develop and evaluate a flexible regression-based approach for heritability inference.
• To accommodate both single and multiple trait analyses, including non-genetic and non-additive genetic factors.

Main Methods:

• Employed regression models for the correlation parameter to estimate heritability.
• Compared the proposed regression method with the likelihood approach.
• Utilized simulations and the Northern Manhattan Family Study (carotid Intima Media Thickness data) for performance evaluation.

Main Results:

• The regression approach demonstrated reliable heritability estimation.
• The proposed methods showed flexibility in incorporating diverse genetic and non-genetic information.
• Performance was comparable to the established likelihood approach.

Conclusions:

• Regression models offer a robust and adaptable tool for heritability estimation.
• This approach enhances the planning and execution of genetic studies for trait-gene discovery.