Including non-informative parents in transmission-based association tests

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• The classical transmission disequilibrium test (TDT) primarily uses data from heterozygous parents, overlooking information from homozygous parents.
• Offspring phenotypes are influenced by alleles from both parents, regardless of parental homozygosity, offering valuable association data.

Purpose of the Study:

• To propose modified TDT procedures that incorporate transmission data from both parents in informative trios.
• To assess the potential gain in statistical power for detecting genetic associations.

Main Methods:

• Development of a goodness-of-fit chi-squared test for binary traits.
• Devising a bivariate response logistic model and a quasi-likelihood approach for quantitative traits.
• Extensive simulations to evaluate type 1 errors and power compared to classical TDT.

Main Results:

• The modified logistic regression approach showed marginally increased power.
• The quasi-likelihood approach demonstrated a substantial gain in power.
• Proposed methods were applied to analyze an alcoholism-related count phenotype.

Conclusions:

• Incorporating transmission data from non-informative parents enhances the power of TDT, particularly with the quasi-likelihood method.
• These modified TDT procedures offer improved genetic association detection for various trait types.