Abstract

DNA methylation is an indispensable epigenetic modification required for regulating the expression of mammalian genomes. Continued efforts have been made to unravel the methylation states genome-wide, featuring the methyl-DNA immunoprecipitation(MeDIP) coupled with next-generation sequencing. Our method uses a single-CpG-resolution, whole-genome methylation caller designed specifically for MeDIP-seq data. It did not require external database for copy number adjustment. Furthermore, it effectively detected genomic regions potentially predisposed to oncogenesis through its prediction of methylation states. The above suggests that our method makes a handy and reliable tool to generate genome-wide methylation profiles. All source codes in PERL language are available upon request of the first author.

Key words: Methyl-DNA immunoprecipitation, next-generation sequencing, Hidden Markov chain.