SASM-VGWAS

SASM-VGWAS integrates spatial-anatomical similarity into voxel-wise genome-wide association studies (VGWAS) to detect imaging and genetic biomarkers associated with Alzheimer's disease.

Key Features:

• Spatial-Anatomical Similarity Matrix: Captures correlations among voxels in brain imaging data to represent spatial relationships within anatomically meaningful regions.
• Modified Simple Linear Iterative Clustering: Employs a tailored modification of the simple linear iterative clustering method to achieve spatial grouping reflecting characteristics of Alzheimer's disease.
• Efficient VGWAS Framework: Incorporates spatial correlations into the VGWAS analysis to reduce computational demands while improving biomarker detection accuracy.
• Imaging and Genetic Data Integration: Combines imaging and genetic data to identify putative AD biomarkers and associations between brain regions and genetic variation.
• Validation on ADNI Dataset: Methodology was validated on data from 708 subjects from the Alzheimer's Disease Neuroimaging Initiative to detect risk genes and voxel clusters associated with AD.

Scientific Applications:

• Imaging Genomics Biomarker Discovery: Detects imaging and genetic biomarkers relevant to neurodegenerative diseases, particularly Alzheimer's disease.
• Subject Classification: Uses detected biomarkers as predictors to classify subjects into AD versus normal control groups.
• Risk Gene and Cluster Identification: Facilitates discovery of new risk genes and spatial voxel clusters associated with Alzheimer's disease pathology.

Methodology:

Constructs a spatial-anatomical similarity matrix, applies a modified simple linear iterative clustering method for spatial grouping, incorporates spatial correlations into a VGWAS framework, integrates imaging and genetic data, and validates results on the ADNI dataset of 708 subjects.