Skip to main navigation Skip to search Skip to main content

CpG_MI: A novel approach for identifying functional CpG islands in mammalian genomes

  • Jianzhong Su
  • , Yan Zhang*
  • , Jie Lv
  • , Hongbo Liu
  • , Xiaoyan Tang
  • , Fang Wang
  • , Yunfeng Qi
  • , Yujia Feng
  • , Xia Li
  • *Corresponding author for this work
  • Harbin Medical University

Research output: Contribution to journalArticlepeer-review

Abstract

CpG islands (CGIs) are CpG-rich regions compared to CpG-depleted bulk DNA of mammalian genomes and are generally regarded as the epigenetic regulatory regions in association with unmethylation, promoter activity and histone modifications. Accurate identification of CpG islands with epigenetic regulatory function in bulk genomes is of wide interest. Here, the common features of functional CGIs are identified using an average mutual information method to differentiate functional CGIs from the remaining CGIs. A new approach (CpG mutual information, CpG_MI) was further explored to identify functional CGIs based on the cumulative mutual information of physical distances between two neighboring CpGs. Compared to current approaches, CpG_MI achieved the highest prediction accuracy. This approach also identified new functional CGIs overlapping with gene promoter regions which were missed by other algorithms. Nearly all CGIs identified by CpG_MI overlapped with histone modification marks. CpG_MI could also be used to identify potential functional CGIs in other mammalian genomes, as the CpG dinucleotide contents and cumulative mutual information distributions are almost the same among six mammalian genomes in our analysis. It is a reliable quantitative tool for the identification of functional CGIs from bulk genomes and helps in understanding the relationships between genomic functional elements and epigenomic modifications.

Original languageEnglish
Article numbergkp882
Pages (from-to)e6-e6
JournalNucleic Acids Research
Volume38
Issue number1
DOIs
StatePublished - 23 Oct 2009
Externally publishedYes

Fingerprint

Dive into the research topics of 'CpG_MI: A novel approach for identifying functional CpG islands in mammalian genomes'. Together they form a unique fingerprint.

Cite this