PePr

PePr prioritizes consistent and differential ChIP-Seq peaks across biological replicates by modeling genome-wide read counts with a negative binomial distribution and applying local variance estimation.

Key Features:

• Biological Replicates Analysis: Handles ChIP-Seq datasets with biological replicates to detect consistent and differential signals for transcription factors (TFs) and histone modifications.
• Negative Binomial Distribution Modeling: Models genome-wide read counts using a negative binomial distribution suitable for sequencing count data.
• Local Variance Estimation: Employs local variance estimation to prioritize binding sites with stable signal across replicates over highly variable regions.
• False Discovery Rate (FDR) Scaling: Provides improved scaling of False Discovery Rate (FDR) estimation for differential region detection.

Scientific Applications:

• Transcription Factor Binding Sites: Identifies genome-wide DNA-binding sites for transcription factors by detecting enriched read-count regions often associated with high motif occurrence rates.
• Histone Modification Analysis: Detects differential and consistently enriched broad regions in histone modification ChIP-Seq data and supports FDR-controlled comparisons between groups.

Methodology:

Models genome-wide read counts with a negative binomial distribution and applies local variance estimation to prioritize consistent versus differential binding sites.