Family-based association test for time-to-onset data with time-dependent differences between the hazard functions.

Area of Science:

• Genetics
• Biostatistics
• Epidemiology

Background:

• Genetic association studies often encounter time-dependent differences in hazard functions between genotypes.
• Non-proportional hazards present analytical challenges in traditional genetic association methods.

Purpose of the Study:

• To introduce a novel weighted Family-Based Association Test (FBAT)-Logrank approach for analyzing genetic data with non-proportional hazards.
• To maximize the statistical power of genetic association tests by optimizing weighting schemes.

Main Methods:

• Utilized the Fleming-Harrington weighted logrank approach.
• Developed a non-parametric estimator for the genetic marker distribution under the alternative hypothesis.
• Integrated estimated marker distributions to determine optimal weights for the FBAT-Logrank test.

Main Results:

• The weighted FBAT-Logrank test demonstrated significant power increases (up to 100%) compared to the original FBAT-Logrank test.
• The methodology effectively handles non-proportional hazards in genetic association studies.
• Analysis provided insights into the age-specific influence of genotypes on disease risk.

Conclusions:

• The proposed weighted FBAT-Logrank method is a powerful tool for genetic association studies with time-dependent effects.
• This approach enhances the ability to detect genotype-disease associations and understand their temporal dynamics.
• The methodology has practical relevance, as shown in an asthma study application.