Robust estimation of gene frequency and association parameters

Area of Science:

• Statistical Genetics
• Biostatistics
• Genomic Data Analysis

Background:

• Accurate estimation of gene frequency and association parameters is crucial for understanding genetic diseases.
• Traditional methods may be sensitive to model assumptions and data structure in family studies.
• Robust statistical methods are needed to handle potential misspecification in genetic models.

Purpose of the Study:

• To propose novel methods for estimating and testing gene frequency and association parameters using family data.
• To develop statistically robust procedures that are less sensitive to model assumptions.
• To evaluate the efficiency of the proposed methods compared to existing techniques.

Main Methods:

• Utilizing generalized estimating equations (GEE) for parameter estimation.
• Employing marginal models to estimate gene frequency and association parameters within individual distributions.
• Developing procedures for allele frequency estimation, linkage analysis, Hardy-Weinberg disequilibrium testing, and marker-disease association studies.

Main Results:

• The proposed GEE-based methods provide consistent estimates of gene frequency and association parameters and their standard errors.
• These methods demonstrate robustness, yielding reliable results even when the marginal model is not perfectly specified.
• The methods exhibit good efficiency when compared to maximum likelihood estimation techniques.

Conclusions:

• Generalized estimating equations offer a robust and efficient framework for analyzing genetic parameters in family studies.
• The proposed methods are applicable to a range of genetic analyses, including allele frequency estimation and association testing.
• These techniques provide reliable statistical inference for genetic data, even under model uncertainty.