Bayesian nonparametric method for genetic dissection of brain activation region

Area of Science:

• Neuroimaging Genetics
• Statistical Bioinformatics
• Computational Neuroscience

Background:

• Brain atrophy is linked to Alzheimer's disease (AD) neuropathology.
• Formal statistical methods for dissecting brain activation phenotypes (shape, intensity) are lacking.
• Understanding genetic influences on brain changes is critical for AD research.

Purpose of the Study:

• To develop a statistical framework for the genetic dissection of brain activation phenotypes in Alzheimer's disease.
• To integrate shape and intensity analysis of brain activation regions.
• To identify genetic variants associated with brain activation intensity.

Main Methods:

• Proposed a two-level Bayesian hierarchical model.
• Level 1: Bayesian nonparametric level set (BNLS) model for brain activation region shape.
• Level 2: Regression model with spike-and-slab and Gaussian priors for genetic variant selection based on brain activation intensity.
• Utilized Markov chain Monte Carlo (MCMC) for posterior computation.

Main Results:

• The proposed Bayesian hierarchical model effectively analyzes brain activation phenotypes.
• The method successfully identifies genetic variants associated with brain activation intensity.
• Demonstrated advantages through simulation studies and analysis of Alzheimer's Disease Neuroimaging Initiative (ADNI) data.

Conclusions:

• The developed Bayesian model provides a robust statistical approach for imaging genetics in Alzheimer's disease.
• This method facilitates the genetic dissection of complex brain activation phenotypes.
• Findings contribute to a deeper understanding of the genetic underpinnings of Alzheimer's disease.