BaalChIP

BaalChIP models allele-specific binding from multiple ChIP-seq BAM files by computing allele counts at individual variants and using a Bayesian framework to correct for background allele frequency and copy-number variation to detect allele-specific events.

Key Features:

• Allele-Specific Event Detection: Computes allele counts at individual variants, including single nucleotide polymorphisms (SNPs) and single nucleotide variants (SNVs), to detect allele-specific transcription factor binding.
• Quality Control Measures: Implements extensive quality-control steps to filter out problematic variants prior to allele-specific analysis.
• Bayesian Framework: Applies a Bayesian statistical approach to correct for the effect of background allele frequency on observed ChIP-seq read counts and to account for copy-number variation (CNV).
• Joint Analysis Across Samples: Performs joint analysis of multiple ChIP-seq samples across a single variant to increase statistical power for allele-specific detection.
• Enhanced Detection Power for Cis-Regulatory Variants: Correcting for CNVs increases sensitivity to identify putative cis-acting regulatory variants in cancer genomes.

Scientific Applications:

• Allelic effects of non-coding variants: Provides allele-specific measurements of transcription factor binding to study the impact of non-coding variants on regulatory activity.
• Cancer genomics: Handles samples with prevalent copy-number variation to enable accurate allele-specific analyses in cancer genomes.
• Regulatory variant identification: Enhances detection of putative cis-acting regulatory variants that may contribute to phenotypic diversity and disease.
• Dissection of regulatory mechanisms: Aids investigation of how genetic variation affects transcription factor binding and regulatory architecture.

Methodology:

Computes allele counts from ChIP-seq BAM files, applies quality-control filtering of variants, uses a Bayesian statistical framework to correct for background allele frequency and CNVs, supports joint analysis across multiple samples, and has been validated on 548 ENCODE ChIP-seq and six targeted FAIRE-seq samples.