TAGOOS

TAGOOS prioritizes non-coding loci associated with complex phenotypes by integrating GWAS-associated SNPs, linkage disequilibrium (LD) blocks, and regulatory molecular markers into a supervised-learning framework to compute genome-wide functional scores.

Key Features:

• Supervised learning: Employs genome-wide supervised learning to distinguish putative regulatory SNPs in non-coding regions.
• LD integration: Integrates associated SNPs with linkage disequilibrium (LD) blocks to account for LD structure around signals.
• Regulatory marker integration: Incorporates molecular markers indicative of gene regulatory features to inform prioritization.
• Genome-wide scoring: Computes genome-wide scores that enrich and prioritize previously unobserved associated SNPs.
• Performance metrics: Reports enrichment with odds ratios of 4.3 for intronic regions and 3.5 for intergenic regions and AUCs of 0.65 (intronic) and 0.60 (intergenic).
• Correlation with genomic indicators: Scores correlate with maximal SNP significance, expression quantitative trait loci (eQTLs), and the number of biological samples annotated with regulatory features.
• Locus recovery and prediction: Recovers known functional loci and predicts novel loci enriched with transcription factors relevant to studied phenotypes.
• Outputs: Produces numeric scores, annotations, and UCSC genome tracks for downstream analysis.

Scientific Applications:

• GWAS prioritization: Prioritizes putative functional non-coding SNPs and loci identified by genome-wide association studies.
• Regulatory element identification: Identifies candidate regulatory elements linked to complex traits through integration with eQTLs and regulatory annotations.
• Trait-specific discovery: Applied to phenotypes such as cleft lip and human adult height to recover known loci and predict novel, TF-enriched loci.

Methodology:

Integrates GWAS-associated SNPs with LD blocks and regulatory molecular markers into a supervised learning model to compute genome-wide enrichment/prioritization scores.