ChIP-seq is a powerful technology for detecting genomic regions where a protein of interest interacts with DNA. ChIP-seq data for mapping transcription factor binding sites (TFBSs) have a characteristic pattern: around each binding site, sequence reads aligned to the forward and reverse strands of the reference genome form two separate peaks shifted away from each other, and the true binding site is located in between these two peaks. While it has been shown previously that the accuracy and resolution of binding site detection can be improved by modeling the pattern, efficient methods are unavailable to fully utilize that information in TFBS detection procedure. We present PolyaPeak, a new method to improve TFBS detection by incorporating the peak shape information. PolyaPeak describes peak shapes using a flexible PĆ³lya model. The shapes are automatically learnt from the data using Minorization-Maximization (MM) algorithm, then integrated with the read count information via a hierarchical model to distinguish true binding sites from background noises. Extensive real data analyses show that PolyaPeak is capable of robustly improving TFBS detection compared with existing methods. An R package is freely available.


  1. PolyaPeak: detecting transcription factor binding sites from ChIP-seq using peak shape information.
    Cite this
    Wu H, Ji H, 2014-01-01 - PloS one


  1. Hao Wu

    Department of Biostatistics and Bioinformatics, Emory University, United States of America

  2. Hongkai Ji

    Department of Biostatistics, Johns Hopkins University, United States of America

Community Ratings

UsabilityEfficiencyReliabilityRated By
0 user
Sign in to rate
Tool TypeApplication
User InterfaceTerminal Command Line
Download Count0
Country/RegionUnited States of America
Submitted ByHongkai Ji