A statistical method for image-mediated association studies discovers genes and pathways associated with four brain...

Area of Science:

• Neuroscience
• Genetics
• Medical Imaging

Background:

• Brain imaging and genomics are essential for understanding brain disorder genetics.
• Acquiring imaging data for large cohorts is costly and often not feasible for historical genome-wide association studies (GWAS) datasets.
• Legacy GWAS datasets lack crucial neuroimaging information for integrated genetic analysis.

Purpose of the Study:

• To develop a cost-effective method for identifying the genetic underpinnings of brain disorders in legacy GWAS cohorts.
• To leverage existing neuroimaging data to enable association mapping in datasets lacking such information.
• To demonstrate the power of an integrated imaging-genomics approach for discovering genetic risk factors for neuropsychiatric disorders.

Main Methods:

• Developed an image-mediated association study (IMAS) using an integrated feature selection/aggregation model.
• Utilized UK Biobank image-derived phenotypes (IDPs) as borrowed data for association mapping.
• Validated discovered genetic associations through analysis of annotations, pathways, and expression quantitative trait loci (eQTLs).

Main Results:

• Successfully identified genetic bases for four neuropsychiatric disorders using the IMAS approach.
• Discovered a common cerebellar-mediated mechanism underlying the four studied neuropsychiatric disorders.
• Simulations indicated IMAS is more powerful than hypothetical protocols using imaging data directly within GWAS datasets for identifying genetic risk.

Conclusions:

• The developed IMAS method enables effective association mapping in legacy GWAS cohorts by utilizing external imaging data.
• IMAS offers a cost-effective strategy for reanalyzing existing GWAS datasets, facilitating integrated genetics and imaging research without new imaging acquisition.
• This approach significantly enhances the potential for discovering genetic risk factors for brain disorders by bridging legacy genetic data with available imaging phenotypes.