Computational analysis of genome-wide DNA methylation during the differentiation of human embryonic stem cells along the endodermal lineage

Abstract

The generation of genome-wide data derived from methylated DNA immunoprecipitation followed by sequencing (MeDIP-seq) has become a major tool for epigenetic studies in health and disease. The computational analysis of such data, however, still falls short on accuracy, sensitivity, and speed. We propose a time-efficient statistical method that is able to cope with the inherent complexity of MeDIP-seq data with similar performance compared with existing methods. In order to demonstrate the computational approach, we have analyzed alterations in DNA methylation during the differentiation of human embryonic stem cells (hESCs) to definitive endoderm. We show improved correlation of normalized MeDIP-seq data in comparison to available whole-genome bisulfite sequencing data, and investigated the effect of differential methylation on gene expression. Furthermore, we analyzed the interplay between DNA methylation, histone modifications, and transcription factor binding and show that in contrast to de novo methylation, demethylation is mainly associated with regions of low CpG densities.[Supplemental material is available at http://www.genome.org. The MeDIP-seq data from this study have been submitted to NCBI Sequence Read Archive (http://www.ncbi.nlm.nih.gov/Traces/sra/sra.cgi) under accession no. SRA012665. The bead array gene expression data from this study have been submitted to the NCBI Gene Expression Omnibus (http:// www.ncbi.nlm.nih.gov/geo) under accession no. GSE21715. The MEDIPS software package, manual, data, and example data are available online at http://medips.molgen.mpg. et al. 2005), which can be detected either by tiling arrays (MeDIPChip) or by next-generation sequencing (MeDIP-seq). Methylation profiles obtained by the MeDIP approach are not base pair-specific but reflect methylation levels on a resolution restricted by the size of the sonicated DNA fragments after amplification and size selection. In contrast, bisulfite sequencing detects cytosine methylation on a base-pair level. Although whole-genome single-base resolution maps have been generated (Lister et al. 2008(Lister et al. , 2009, such techniques cannot yet be cost-effectively applied to screen large sets of sequences or samples. Reduced representation bisulfite sequencing (RRBS) was introduced in order to address this issue by selecting only some regions of the genome for sequencing. Here, reduced representation is achieved by the size-fractionation of DNA fragments after restriction enzyme digestion (Meissner et al. 2008;Laird 2010). In contrast to bisulfite sequencing, MeDIP-seq-derived methylation data are of far lower resolution, and therefore, it remains difficult to discriminate between CpG and non-CpG methylation when single-end short reads are considered. However, MeDIP-seq covers nearly as many CpGs per sample genome as does the more expensive whole-genome shotgun bisulfite sequencing (WGSBS) approach (Laird 2010). An advantage of the MeDIP approach is the generation of unbiased, cost-effective, and fullgenome methylation levels w...

Epigenome-wide profiling of DNA methylation in paired samples of adipose tissue and blood

“…38 There is cell mixture even within adipose tissue, and genes are differentially expressed in different adipocytes 39 or within the same cell type. 40 A statistical method has been proposed to adjust for latent classes of DNA methylation attributable to cell mixture. 41 Hence, accounting for cellular heterogeneity within individual tissues is an important future direction for this research.…”

Tcl1 protein functions as an inhibitor of de novo DNA methylation in B-cell chronic lymphocytic leukemia (CLL)

“…Sequence reads were analyzed using the Illumina ELAND primary analysis pipeline and subsequently aligned to the mm9 mouse genome using Bowtie software, allowing up to three mismatches. The aligned sequences were analyzed with MEDIPS software (25). The criteria for region selection were (i) ≥0.15 reads per million in at least one of the two comparisons; (ii) P ≤ 0.05 in both Wilcoxon and Student t tests; and (iii) a twofold to fourfold or greater change in reads per million.…”

Genome-wide DNA Methylation Events in TMPRSS2–ERG Fusion-Negative Prostate Cancers Implicate an EZH2-Dependent Mechanism with miR-26a Hypermethylation

“…However, the epigenetic landscapes of prostate cancer subgroups are still not suffi ciently investigated to draw any conclusions. To understand alterations in the ERG fusion-negative class of prostate cancer, we used a deep sequencing readout of methylated DNA immunoprecipitation sequencing (MeDIPSeq) to screen 51 tumor and 53 benign prostate tissues (23)(24)(25)(26). We integrated the results with gene and microRNA (miRNA) expression analyses and proposed a model for the development of aberrant DNA methylation patterns in ERG fusion-negative (FUS − ) prostate cancers.…”