Fitting Procedures for Novel Gene-by-Measured Environment Interaction Models in Behavior Genetic Designs

Area of Science:

• Behavioral Genetics
• Quantitative Genetics
• Biometrics

Background:

• Quantitative behavior genetic studies often examine gene-by-environment interactions (GxM).
• Existing models by Purcell and Rathouz et al. address GxM interactions and gene-environment correlation but have limitations.
• Non-linear biometric models offer a broader framework for these analyses.

Purpose of the Study:

• To propose a novel factorization of the likelihood for non-linear biometric models.
• To adopt numerical integration techniques for model estimation, particularly for models lacking closed-form likelihoods.
• To establish the validity of these new procedures through simulation and application.

Main Methods:

• Developed a novel factorization of the likelihood function for a class of non-linear biometric models.
• Employed numerical integration techniques for model estimation.
• Validated the proposed methods using numerical simulation studies and applied them to a twin study.

Main Results:

• The proposed factorization and numerical integration techniques enable the estimation of complex non-linear biometric models.
• Simulation studies confirmed the validity and accuracy of the new procedures.
• The methods were successfully applied to a twin study, demonstrating the moderating effect of birth weight on genetic influences on childhood anxiety.

Conclusions:

• The novel factorization and numerical integration provide a flexible and robust approach for analyzing gene-environment interactions in quantitative genetics.
• These methods extend the applicability of biometric models to complex, non-linear relationships.
• The findings have implications for understanding the interplay of genetic and environmental factors in complex traits like childhood anxiety.