Modeling maternal-offspring gene-gene interactions: the extended-MFG test

Area of Science:

• Genetics
• Biostatistics
• Reproductive Health

Background:

• Maternal-fetal genotype (MFG) incompatibility, where maternal and offspring genes interact negatively, increases fetal disease susceptibility.
• Existing statistical methods for MFG incompatibility are limited to nuclear families, posing challenges for complex, multi-generational family structures.
• Complex family structures in genetic studies can hinder accurate analysis of disease risk factors.

Purpose of the Study:

• To develop a novel statistical method, the Extended-MFG (EMFG) Test, for analyzing MFG incompatibility in arbitrary family structures.
• To adapt existing nuclear-family based MFG tests for complex pedigrees, enhancing their applicability and power.
• To improve the accuracy of parameter estimation and statistical power in MFG incompatibility studies, particularly when genetic data is incomplete.

Main Methods:

• Developed the Extended-MFG (EMFG) Test, a model-based likelihood approach utilizing Ott's pedigree likelihood representation.
• Modified the nuclear-family "mating type" approach to accommodate complex family structures and arbitrary mating.
• Integrated calculation of MFG incompatibility with Mendelian transmission probabilities, assuming random mating at the locus of interest.

Main Results:

• Simulations demonstrate that the EMFG Test maintains appropriate type-I error rates and exhibits suitable power for detecting MFG incompatibility.
• The EMFG Test provides precise parameter estimation when the assumption of random mating holds.
• Simulations and real data examples show improved accuracy in parameter estimation and increased power compared to the nuclear-family MFG test, especially with missing genotypes.

Conclusions:

• The Extended-MFG (EMFG) Test offers a robust and flexible statistical framework for investigating maternal-fetal genotype incompatibility in diverse and complex family structures.
• The EMFG Test enhances the ability to identify genetic interactions contributing to disease risk, overcoming limitations of previous methods.
• This approach is particularly advantageous for large-scale genetic studies where family structures are intricate and genotype data may be incomplete, leading to more accurate disease risk assessments.