ZINBA

ZINBA identifies enriched genomic regions from ChIP-seq and DNA-seq next-generation sequencing data using a zero-inflated negative binomial statistical framework to distinguish true signal from background.

Key Features:

• Zero-Inflated Negative Binomial modeling: Implements a zero-inflated negative binomial statistical model to represent count data with excess zeros.
• Broad and narrow enrichment detection: Detects both broad and narrow modes of enrichment across genomic regions.
• Supports ChIP-seq and DNA-seq: Applicable to ChIP-seq and related next-generation sequencing experiments, including DNA-seq.
• Covariate modeling (G/C content): Models and accounts for factors that co-vary with background or experimental signals such as G/C content.
• Local copy number variation handling: Accounts for complex genomic contexts including regions with local copy number variations.
• Robust across signal-to-noise ratios: Operates effectively across diverse signal-to-noise ratios.
• Unified framework for DNA-seq analysis: Provides a unified statistical framework for analyzing DNA-seq experiments.

Scientific Applications:

• ChIP-seq enrichment detection: Identifying enriched genomic regions in ChIP-seq experiments.
• DNA-seq enrichment analysis: Identifying enriched regions in DNA-seq experiments.
• Analysis in complex genomic landscapes: Detecting enrichment in regions with local copy number variation or other complex genomic contexts.
• Analysis across variable signal-to-noise: Detecting both broad and narrow enrichment across datasets with diverse signal-to-noise ratios.

Methodology:

Uses a zero-inflated negative binomial statistical model with explicit covariate modeling (e.g., G/C content) and accounts for local copy number variation while detecting broad and narrow enrichment across varying signal-to-noise ratios.