Illumina’s MethylationEPIC array interrogates the methylation status of over 850,000 cytosines in the human genome. Since Because the MethylationEPIC array is closely related to the Infinium HumanMethylation450 BeadChip, the steps presented in this document can be applied to either platform. The analysis laid out in the tutorial is based on detection of differentially methylated CpG sites and their subsequent filtering and annotation.

• Importing and normalizing methylation data
• Annotating the samples
• Performing exploratory data analysis
• Detecting differentially methylated loci
• Obtaining methylation signatures
• Biological interpretation
• Detecting differentially methylated CpG islands regions

Due to annotation changes, the results presented here may slightly differ over the time.

This tutorial illustrates how to:

The data set accompanying this document consists of eight sixteen human samples processed by Illumina MethylationEPIC  BeadChip arrays. The data set is taken from a study of DNA methylation in human pluripotent stem cells (hPCS) (Lee et al. Cell Rep 2017).

Naive hPSCs can be generated and cultured in vitro from somatic cells. Unfortunately, while naive-state hPSCs have robust self-renewal ability and can be maintained in cell culture, they can only differentiate into the neural fate in vitro. Multi-germ layer potential can be reached by converting naive-state hPSCs to primed-state hPSCs. Unfortunately, this process requires long-term serial passage of naive-state hPSCs under priming conditions. A faster and simpler method to expand the in vitro differentiation potential of naive hPSCs would increase their utility.  The authors hypothesized that the higher expression of the core pluripotent genes OCT4 and NANOG in naive vs. primed hPSCs is responsible for the reduced differentiation capacity of naive hPSCs by enhancing self-renewal at the expense of differentiation capacity. To test this hypothesis, the authors reduced expression of either both OCT4 and NANOG using shRNAs against their common regulator POU5F1 or only NANOG using shRNAs. DNA methylation patterns serve as markers for cell state and were compared between native hPSCs, primed hPSCs, native hPSCs + POU5F1-shRNAs, and native hPSCs + NANOG-shRNAs. The results of this analysis indicate that OCT4 levels play an important role in regulating differentiation potential vs. self-renewal of naive hPSCs. 

B cells and  B cells infected with Epstein-Barr virus (EBV). 

Infecting B cells with EBV in vitro transforms them, making them capable of indefinite growth in vitro. These immortalized cell lines are referred to as lymphoblastoid cell lines (LCLs). LCLs behave similarly to activated B cells, making them useful for expanding T cells in vitro. Because EBV is a carcinogen and immortalized cell growth is a hallmark of cancer, examining the effects of EBV transformation on B cell DNA methylation might shed light on the roles of DNA methylation in tumor development.  

The data files can be downloaded from Gene Expression Omnibus using accession number GSE95531number GSE93373 or by selecting this link - Differential Methylation Analysis data set. To follow this tutorial, download the 16 32 .idat files associated with GSE95531 (note that two .idat files are generated for each array) and unzip them on your local computer . using 7-zip, WinRAR, or a similar program.