bootRanges

bootRanges generates null sets of genomic ranges for hypothesis testing by using block bootstrap to preserve local genomic correlation structures and improve accuracy in enrichment and related analyses.

Key Features:

• Block Bootstrapping Methodology: Employs block bootstrap to generate null genomic-range sets while maintaining local correlation structures.
• Enhanced Null Distribution Accuracy: Produces broader and more reliable null distributions for test statistics, reducing inflated significance from simple shuffling or permutation.
• Integration with Bioconductor: Integrates with R/Bioconductor via the nullranges package to compute test statistics within modular analysis workflows.
• Versatility in Genomic Analysis: Applicable to enrichment analysis, cis-regulatory element (CRE)-gene correlation studies across cell types, and optimization of thresholds such as log fold change (logFC) from differential expression analyses.

Scientific Applications:

• Enrichment Analysis: Determines statistically significant associations between genomic features using null models that preserve local correlation.
• Complex Genomic Studies: Supports analysis of CRE–gene correlations across cell types and threshold optimization (e.g., logFC) in differential expression studies.

Methodology:

Applies block bootstrap to construct null genomic-range sets that preserve local correlation structures, contrasts this approach with traditional shuffling or permutation methods, and is implemented within the R/Bioconductor nullranges framework.