Computational methods for epigenetic analysis: the protocol of computational analysis for modified methylation-specific digital karyotyping based on massively parallel sequencing.

Massively parallel sequencing technology opens new possibilities for epigenetic research. Many methods have been developed based on the new sequencing platforms, allowing an ultra-deep mapping of epigenetic variants in a fast and cost-effective way. However, handling millions of short reads produced by these sequencing platforms is a huge challenge for many laboratories. Thus, there is a need for the development of accurate and fast computational tools for epigenetic studies in the new era of genomic sequencing.Modified methylation-specific digital karyotyping (MMSDK) is an improved method for genome-wide DNA methylation profiling based on the combination of traditional MSDK and Illumina/Solexa sequencing. Here, we introduce our computational tools used in the MMSDK analysis process from the experimental design to statistical analysis. We have developed a mapping process based on the in silico simulation of combined enzyme cutting and tag extraction of the reference genome. Subsequently, the 20-21 nucleotides (nt) long tags obtained by sequencing are mapped to the simulated library using an open source software Mapping and Assembly with Qualities. Our computational methods include trimming, annotation, normalization, and counting the reads to obtain digital DNA methylation profiles. We present the complete protocol and discuss some important issues that should be considered by readers, such as handling of repeat sequences, SNPs, and normalization. The core part of this protocol (mapping and annotation of tags) is suitable for any tag profiling-based methods, and it could also be modified to analyze results from other types of epigenetic studies based on massively parallel sequencing.