Associating Multi-Modal Brain Imaging Phenotypes and Genetic Risk Factors via a Dirty Multi-Task Learning Method

Area of Science:

• Neuroscience
• Genetics
• Medical Imaging

Background:

• Brain imaging genetics is crucial for understanding brain disorders.
• Multi-modal imaging data may offer complementary information but shared variance is unclear.
• Complex genetic mechanisms underlying brain structure/function require advanced analytical methods.

Purpose of the Study:

• To propose a novel dirty multi-task sparse canonical correlation analysis (SCCA) for multi-modal brain imaging genetics.
• To identify shared and modality-specific imaging quantitative traits (QTs) and genetic loci.
• To analyze complex associations between genetic variations and brain phenotypes across multiple imaging modalities.

Main Methods:

• Developed a dirty multi-task sparse canonical correlation analysis (SCCA) integrating multi-task learning and parameter decomposition.
• Applied the method to multi-modal brain imaging quantitative traits (QTs).
• Compared performance against state-of-the-art multi-view SCCA and multi-task SCCA using synthetic and real neuroimaging genetic data.

Main Results:

• The proposed dirty multi-task SCCA demonstrated superior or comparable canonical correlation coefficients and weights.
• Successfully identified both shared and modality-specific imaging QTs and genetic loci.
• Revealed complex multi-SNP-multi-QT associations across different brain imaging modalities.

Conclusions:

• The dirty multi-task SCCA is a powerful and flexible tool for multi-modal brain imaging genetics.
• It provides meaningful insights into modality-consistent and modality-specific biomarkers.
• This method advances the study of genetic influences on brain structure and function.