yarn

yarn performs large-scale RNA-Seq processing and analysis, providing mis-annotation quality control, filtering, condition-aware normalization, and cross-tissue cis- and trans-eQTL discovery to investigate genetic regulation of gene expression.

Key Features:

• Mis-annotation quality control: Detects and flags sample mis-annotations in RNA-Seq datasets.
• Filtering: Applies gene- and sample-level filtering to address sparsity in large RNA-Seq experiments.
• Condition-aware normalization: Performs normalization that accounts for experimental conditions and tissue heterogeneity.
• Bioconductor integration: Leverages existing Bioconductor tools and statistical methods for downstream analyses.
• Cross-tissue eQTL analysis: Identifies cis- and trans-eQTLs across multiple tissues, exemplified by analyses of GTEx v6.0 across 13 tissues.
• Bipartite SNP–gene network construction: Builds tissue-specific bipartite networks where significant SNP–gene associations are represented as edges.
• Community detection: Reveals dense, highly modular communities within SNP–gene networks that correspond to coherent biological processes.
• Global network hub characterization: Identifies global hubs enriched in distal regulatory regions such as enhancers and depleted for GWAS-associated SNPs.
• Core SNP (local hub) identification: Detects community-specific core SNPs located in promoters and enhancers that are enriched for trait and disease associations.

Scientific Applications:

• eQTL discovery across tissues: Mapping cis- and trans-acting genetic variants that influence gene expression in multiple tissues.
• Regulatory variant prioritization: Prioritizing SNPs in promoters and enhancers that are likely to affect gene regulation and traits.
• Modular network analysis: Identifying gene–SNP communities corresponding to biological processes for functional interpretation.
• Interpretation of GWAS signals: Assessing enrichment of trait- and disease-associated SNPs within network hubs and communities.

Methodology:

Performs mis-annotation quality control, filtering, and condition-aware normalization; constructs tissue-specific bipartite SNP–gene networks from significant associations; detects dense modular communities; identifies global hubs enriched in enhancers (depleted for GWAS SNPs) and local community-specific core SNPs in promoters/enhancers enriched for trait/disease associations; example analyses use GTEx v6.0 across multiple tissues.