Adaptive testing for multiple traits in a proportional odds model with applications to detect SNP-brain network...

Area of Science:

• Genetics
• Neuroimaging
• Statistical Genetics

Background:

• Detecting genetic associations with multiple traits is crucial in neuroimaging studies.
• Existing methods often assume additive inheritance and treat traits as responses, limiting flexibility.
• Handling mixed trait types (quantitative and binary) and high-dimensional data remains challenging.

Purpose of the Study:

• To develop a powerful and flexible statistical test for detecting genetic associations with multiple traits.
• To propose a proportional odds model (POM) approach treating SNPs as ordinal responses and traits as predictors.
• To create an adaptive test within the POM framework for enhanced power across diverse scenarios.

Main Methods:

• Utilized a proportional odds model (POM) with single nucleotide polymorphisms (SNPs) as ordinal responses and neuroimaging traits as predictors.
• Developed an adaptive statistical test for the POM framework to optimize power.
• Applied the method to analyze structural MRI and resting-state functional MRI (rs-fMRI) data from the Alzheimer's Disease Neuroimaging Initiative (ADNI).

Main Results:

• The proposed POM method effectively handles mixed trait types and is robust to inheritance mode assumptions.
• The adaptive test demonstrates high power across various situations, outperforming existing methods in high-dimensional settings (p>n).
• Analysis of ADNI data identified significant SNPs associated with neuroimaging phenotypes, including structural and functional brain measures.

Conclusions:

• The novel POM-based adaptive test provides a powerful and flexible tool for genetic association studies in neuroimaging.
• This approach is particularly advantageous for high-dimensional genetic data where the number of phenotypes exceeds the sample size.
• The method successfully identified biologically relevant SNPs, highlighting its utility in understanding complex brain disorders.