MOCCS: Clarifying DNA-binding motif ambiguity using ChIP-Seq data. (August 2016)

• Record Type: Journal Article
• Authors: Ozaki, Haruka; Iwasaki, Wataru
• Is Part Of: Computational biology and chemistry. Volume 63(2016)
• Page Start: 62
• Page End: 72
• Publication Date: 2016-08
• Subjects: DNA binding motifs -- ChIP-Seq -- Transcription factors; Chemistry -- Data processing -- Periodicals; Biology -- Data processing -- Periodicals; Biochemistry -- Data processing; Biology -- Data processing; Molecular biology -- Data processing; Periodicals; Electronic journals; 542.85
• DOI: 10.1016/j.compbiolchem.2016.01.014
• Languages: English
• ISSNs: 1476-9271

Abstract

Abstract: Background: As a key mechanism of gene regulation, transcription factors (TFs) bind to DNA by recognizing specific short sequence patterns that are called DNA-binding motifs. A single TF can accept ambiguity within its DNA-binding motifs, which comprise both canonical (typical) and non-canonical motifs. Clarification of such DNA-binding motif ambiguity is crucial for revealing gene regulatory networks and evaluating mutations in cis -regulatory elements. Although chromatin immunoprecipitation sequencing (ChIP-seq) now provides abundant data on the genomic sequences to which a given TF binds, existing motif discovery methods are unable to directly answer whether a given TF can bind to a specific DNA-binding motif. Results: Here, we report a method for clarifying the DNA-binding motif ambiguity, MOCCS. Given ChIP-Seq data of any TF, MOCCS comprehensively analyzes and describes every k -mer to which that TF binds. Analysis of simulated datasets revealed that MOCCS is applicable to various ChIP-Seq datasets, requiring only a few minutes per dataset. Application to the ENCODE ChIP-Seq datasets proved that MOCCS directly evaluates whether a given TF binds to each DNA-binding motif, even if known position weight matrix models do not provide sufficient information on DNA-binding motif ambiguity. Furthermore, users are not required to provide numerous parameters or background genomic sequence models that are typically unavailable. MOCCS is implemented in Perl and R and is freely available viahttps://github.com/yuifu/moccs . Conclusions: By complementing existing motif-discovery software, MOCCS will contribute to the basic understanding of how the genome controls diverse cellular processes via DNA–protein interactions.