ARTP2

ARTP2 performs pathway-based analysis of GWAS data to detect biological pathways enriched for gene-level association signals using SNP-based gene representation and the adaptive rank-truncated product method.

Key Features:

• Pathway-based analysis: Analyzes GWAS data at the pathway level to identify biologically relevant processes beyond individual loci.
• Integration with multiple databases: Leverages pathway definitions from KEGG, BioCarta, and the NCI Protein Interaction Database.
• Gene representation by SNPs: Represents genes by their most strongly associated single nucleotide polymorphisms (SNPs) from GWAS data.
• Enrichment score calculation: Computes enrichment scores using a weighted Kolmogorov-Smirnov procedure to quantify pathway-level overrepresentation of association signals.
• Hierarchical clustering and ARTP methodology: Uses hierarchical clustering to identify pathways with overlapping genes and applies the adaptive rank-truncated product (ARTP) method to detect clusters with excess association signals.
• Implementation: Provided as an R package for executing the described computational analyses.

Scientific Applications:

• Genetic epidemiology: Facilitates interpretation of the collective impact of multiple genetic variants on disease risk through pathway-level analysis.
• Breast cancer GWAS analysis (NCI CGEMS): Has been applied to NCI Cancer Genetic Markers of Susceptibility data to identify enriched pathways such as syndecan-1-mediated signaling, hepatocyte growth factor receptor signaling, growth hormone signaling, and the RAS/RAF/mitogen-activated protein kinase cascade.

Methodology:

Retrieve pathways from KEGG, BioCarta, and the NCI Protein Interaction Database; map each gene to its most strongly associated SNP from GWAS data; calculate enrichment scores using a weighted Kolmogorov-Smirnov test; perform hierarchical clustering of pathways by overlapping genes; apply the adaptive rank-truncated product (ARTP) method to assess clusters for excess association signals.