Statistical equivalent of the classical TDT for quantitative traits and multivariate phenotypes

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Quantitative traits are crucial in clinical research, necessitating advanced statistical methods for association mapping.
• Family-based association tests are preferred for their robustness against population stratification compared to population-based tests.

Purpose of the Study:

• To propose a novel logistic regression model for quantitative trait association testing using a trio design.
• To demonstrate the method's extension from binary trait association tests.
• To evaluate its performance against existing methods like Family-Based Association Test (FBAT).

Main Methods:

• A logistic regression model was developed for quantitative trait association analysis in nuclear families (trios).
• Extensive simulations were conducted to assess the model's statistical power and type I error rate.
• The proposed method was compared with the Family-Based Association Test (FBAT).

Main Results:

• The proposed logistic regression model maintains correct statistical size (avoids false positives) even when non-informative trios are removed, unlike FBAT.
• The method shows comparable power to FBAT when all families are included.
• The model is adaptable for multivariate phenotypes, as demonstrated in an alcoholism endophenotype analysis.

Conclusions:

• The proposed logistic regression model provides a reliable and robust method for quantitative trait association mapping in family studies.
• It offers an advantage over FBAT by maintaining accurate control of false positive rates in specific data scenarios.
• The model's flexibility extends its utility to complex genetic analyses involving multiple quantitative traits.