PSCAN

PSCAN detects spatial clustering of germline variants within protein three-dimensional structures to perform gene-level association testing and identify signal variants associated with complex diseases.

Key Features:

• Protein-Structure-Guided Approach: Uses protein three-dimensional architecture to guide scanning and focus analyses on spatially proximal residues.
• Spatial Clustering Evaluation: Evaluates spatial clustering of germline variants within protein structures to distinguish clustered from dispersed variants.
• Gene-Level Association Testing: Performs gene-level association tests to assess relationships between genes and complex disease phenotypes.
• Signal Variant Detection: Detects signal variants within genes with increased power and specificity relative to existing gene-level methods.
• Hypothesis Generation: Integrates structural information to generate hypotheses about molecular mechanisms and functional domains underlying associations.

Scientific Applications:

• Complex disease gene discovery: Identifying genes associated with complex diseases through structure-informed gene-level association testing.
• Signal localization within proteins: Pinpointing candidate signal variants by mapping association signals onto protein three-dimensional structures.
• Mechanistic inference: Exploring functional domains and molecular mechanisms that may explain observed gene-disease associations.
• Validation and benchmarking: Application and validation using simulation studies and real-data analyses including lipid traits and Alzheimer's disease datasets.

Methodology:

Performs a fast scanning procedure that evaluates spatial clustering of germline variants within protein three-dimensional structures to distinguish disease-causing from benign variants during gene-level association testing.