Gene-Based Association Analysis for Censored Traits Via Fixed Effect Functional Regressions

Area of Science:

• Genetics
• Biostatistics
• Computational Biology

Background:

• Genetic association studies are crucial for understanding disease progression.
• Existing statistical methods for identifying genetic variants affecting survival traits are limited.
• There is a need for robust methods to analyze gene-based survival outcomes.

Purpose of the Study:

• To develop novel statistical methods for gene-based association analysis of survival traits.
• To introduce Cox proportional hazard models incorporating functional regression (FR).
• To evaluate the performance of the proposed methods against existing approaches.

Main Methods:

• Development of fixed-effect Cox proportional hazard models using functional regression (FR).
• Introduction of likelihood ratio test (LRT) statistics for gene-region association analysis.
• Comparison with Burden Test (BT) and Sequence Kernel Association Test (SKAT) via simulations.

Main Results:

• The proposed Cox FR LRT statistics demonstrated well-controlled type I error rates in simulations.
• Cox FR LRT showed comparable or superior power to SKAT across various scenarios.
• Cox FR LRT outperformed Cox BT LRT in power for survival trait association.

Conclusions:

• The developed Cox FR models and LRT statistics provide a powerful tool for gene-based survival trait analysis.
• These methods are applicable to whole genome and whole exome association studies.
• The approach was successfully demonstrated on an age-related macular degeneration dataset.