hierGWAS

hierGWAS implements a hierarchical joint modeling framework to analyze single nucleotide polymorphisms (SNPs) and SNP groups in genome-wide association studies (GWAS) to improve detection of genetic associations.

Key Features:

• Joint Modeling Approach: Employs a multiple generalized linear model to analyze all SNPs jointly, assessing each SNP while accounting for the effects of other SNPs.
• Control of Family-Wise Error Rate (FWER): Maintains control of the family-wise error rate (FWER) to limit false positives in high-dimensional GWAS datasets.
• Data-Driven Refinement: Automatically refines SNP clusters using data-driven criteria to isolate informative subclusters or individual markers.
• Identification of Informative Signals: Performs conditional testing to determine whether a SNP provides additional information about a phenotype beyond other SNPs, distinguishing true associations from spurious correlations.

Scientific Applications:

• High-Dimensional GWAS Analysis: Applied to high-dimensional GWAS datasets to enhance power for detecting genetic associations in complex traits and diseases.
• WTCCC Seven-Disease Analysis: Applied to the Wellcome Trust Case Control Consortium (WTCCC) seven-disease data, identifying SNPs not detected in the original study that were later validated in independent studies.

Methodology:

Uses a multiple generalized linear model for joint modeling of all SNPs, performs conditional testing of SNPs within SNP groups, applies data-driven refinement of SNP clusters, and controls the family-wise error rate (FWER).