chipenrich

chipenrich performs gene set enrichment testing on peaks from Chromatin Immunoprecipitation sequencing (ChIP-seq) experiments by empirically adjusting for confounding factors such as gene locus length to improve biological interpretation.

Key Features:

• Empirical adjustment for gene locus length: Defines gene locus length as the gene body plus surrounding non-coding sequences and corrects for its correlation with peak presence.
• Accommodation of variable peak–length relationships: Models diverse relationships between gene locus length and peak presence observed across datasets, including ENCODE.
• Type I error calibration: Demonstrates well-calibrated type I error on permuted ENCODE ChIP-seq datasets and highlights inflated error or ranking bias in Fisher's exact test and the binomial test implemented in GREAT.
• Biological pattern detection: Identifies distinct enrichment patterns for DNA-binding proteins such as CTCF, JunD, and glucocorticoid receptor α (GRα), including differences related to peak proximity to transcription start sites.

Scientific Applications:

• Gene set enrichment analysis for ChIP-seq: Provides more accurate assessment of gene set enrichment by correcting locus-length confounding.
• Transcription factor function inference: Supports identification of functional roles of DNA-binding proteins such as CTCF, JunD, and GRα.
• Epigenetic and chromatin studies: Applicable to enrichment analyses involving epigenetic modifications and chromatin accessibility.

Methodology:

Performs gene set enrichment testing with an empirical correction for gene locus length (gene body plus surrounding non-coding sequences), evaluates type I error using permuted ENCODE ChIP-seq datasets, and compares performance to Fisher's exact test and the binomial test implemented in GREAT.