OccuPeak: ChIP-Seq peak calling based on internal background modelling.

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• Chromatin immunoprecipitation sequencing (ChIP-seq) is crucial for identifying genome-wide binding sites.
• Existing peak-calling algorithms often rely on input control datasets, which can have biased GC-content.
• This bias can affect the accuracy of background read modeling.

Purpose of the Study:

• To develop a novel peak-calling algorithm that does not require separate input control datasets.
• To improve the accuracy and sensitivity of identifying transcription factor binding and histone modification sites.
• To provide a user-friendly and robust tool for ChIP-seq data analysis.

Main Methods:

• Designed OccuPeak, a peak-calling algorithm utilizing overlapping single reads within ChIP-seq data to model background noise.
• Evaluated OccuPeak's performance against established tools like MACS and CisGenome.
• Assessed peak significance using an 'excess ratio' metric dependent on tag density and global noise levels.

Main Results:

• OccuPeak demonstrated robust model parameters and high sensitivity in enhancer identification.
• It performed comparably to other tools in overlap tests with transcription factor occupancy and regulatory elements.
• Peaks identified by OccuPeak showed significant enrichment with cardiac disease-associated SNPs.

Conclusions:

• OccuPeak effectively models background noise using internal ChIP-seq data, removing the need for input controls.
• The algorithm offers improved sensitivity and user-friendliness for ChIP-seq analysis.
• OccuPeak facilitates straightforward and reliable identification of regulatory elements and associated genetic variants.