- Data Note
- Open Access
Exemplary multiplex bisulfite amplicon data used to demonstrate the utility of Methpat
GigaSciencevolume 4, Article number: 55 (2015)
DNA methylation is a complex epigenetic marker that can be analyzed using a wide variety of methods. Interpretation and visualization of DNA methylation data can mask complexity in terms of methylation status at each CpG site, cellular heterogeneity of samples and allelic DNA methylation patterns within a given DNA strand. Bisulfite sequencing is considered the gold standard, but visualization of massively parallel sequencing results remains a significant challenge.
We created a program called Methpat that facilitates visualization and interpretation of bisulfite sequencing data generated by massively parallel sequencing. To demonstrate this, we performed multiplex PCR that targeted 48 regions of interest across 86 human samples. The regions selected included known gene promoters associated with cancer, repetitive elements, known imprinted regions and mitochondrial genomic sequences. We interrogated a range of samples including human cell lines, primary tumours and primary tissue samples. Methpat generates two forms of output: a tab-delimited text file for each sample that summarizes DNA methylation patterns and their read counts for each amplicon, and a HTML file that summarizes this data visually. Methpat can be used with publicly available whole genome bisulfite sequencing and reduced representation bisulfite sequencing datasets with sufficient read depths.
Using Methpat, complex DNA methylation data derived from massively parallel sequencing can be summarized and visualized for biological interpretation. By accounting for allelic DNA methylation states and their abundance in a sample, Methpat can unmask the complexity of DNA methylation and yield further biological insight in existing datasets.
DNA methylation can be analyzed using a wide range of methods , with bisulfite sequencing considered the current gold standard. Current technologies such as whole genome bisulfite sequencing (WGBS) and reduced representation bisulfite sequencing (RRBS) provide unprecedented detail of methylation patterns throughout the genome, but the complexity of DNA methylation patterns is masked when simple summary metrics are used. For example, most studies of DNA methylation rationalize levels to a percentage value, which typically masks allelic patterns when interpreting the data. We have developed Methpat, a tool that summarizes and visualizes complex DNA methylation data collected by massively parallel sequencing of bisulfite DNA . Using this tool, the DNA methylation state of individual CpG sites and the abundance of allelic patterns can be visualized . Furthermore, by measuring the abundance of allelic DNA methylation patterns, cellular heterogeneity in methylation patterns can now be explored .
The utility of Methpat was demonstrated by measuring DNA methylation in 86 samples (Table 1) across 48 regions of interest (Table 2). This was achieved by using multiplex PCR on bisulfite converted DNA followed by massively parallel sequencing using an Illumina MiSeq Sequencing platform with v3 chemistry. Each sample was indexed and pooled at equimolar concentrations into a single library pool for sequencing. Data has been deposited into GEO with reference identifiers GSE67856  and GSE71804 . A panel of breast cancer cell lines treated with epidermal growth factor and transforming growth factor beta were also analyzed in parallel .
Our initial QC assessment indicated high bisulfite conversion efficiency with very low non-CpG Cs in reads. An additional amplicon that corresponded to a sequence containing no CpG sites was also included as a control, from which all cytosines were observed to have converted to thymidine residues .
The data included here are the Sequence Read Archive files generated from our experiment. These have been aligned onto the hg38 reference genome using Bismark v0.9.0, from which a BAM file for each sample is generated. Using the Bismark_methylation_extractor command, the methylation status of cytosine residues within each read is output to a tab-delimited file. Methpat then operates on this output file to generate both a summarized tab-delimited file of read pattern counts and a HTML file for visualization. We have included the BAM files, Bismark_methylation_extractor output files and Methpat output files as supporting data. Methpat requires a Browser Extensive Data (BED)-format-like file that contains the coordinates for each amplicon of interest, their size and their primer lengths to extract and summarize DNA methylation pattern counts. The flow of data is summarized in Fig. 1.
Our data has the potential to be used to investigate co-methylation , given the unprecedented depth of coverage of the amplicons investigated even in a single MiSeq run. We have interrogated a variety of regions of the genome including repetitive elements and the mitochondrial genome, which remain a challenge for most short read aligners. The interpretation of DNA methylation at repetitive sequence elements has always been a challenge and they are assumed to be methylated . However, the dynamics of repetitive element DNA methylation in cancer  and development  remain areas of interest that can now be properly interpreted with massively parallel sequencing and visualization tools such as Methpat.
Availability of software and requirements
Project name: Methpat
Project home page: http://bjpop.github.io/methpat/
Operating system(s): any POSIX-like operating system (i.e.: Linux, OS X)
Other requirements: Web Browser to view visualization output (HTML file). Suggested browsers include Firefox, Chrome or Safari. Methpat requires output files derived by Bismark (http://www.bioinformatics.babraham.ac.uk/projects/bismark/) and the Bismark_methylation_extractor command. Methpat can be accessed directly from http://bjpop.github.io/methpat/. With further instructions found at the URL.
License: 3-clause BSD License
Any restrictions to use by non-academics: None
A flow diagram of analytical requirements and files can be found in Fig. 1.
Availability of supporting data and materials
BAM files, bismark_methylation_extractor output files and Methpat output files for each sample analyzed in this paper are available in the GigaScience GigaDB repository .
Browser extensible data
Reduced representation bisulfite sequencing
Whole genome bisulfite sequencing
Fraga MF, Esteller M. DNA methylation: a profile of methods and applications. Biotechniques. 2002;33:632–49.
Wong NC, Pope BJ, Candiloro ILM, Korbie D, Trau M, Trau M, et al. MethPat: a tool for the analysis and visualisation of complex methylation patterns obtained by massively parallel sequencing. Submitted.
Mikeska T, Candiloro IL, Dobrovic A. The implications of heterogeneous DNA methylation for the accurate quantification of methylation. Epigenomics. 2010;2:561–73.
Houseman EA, Accomando WP, Koestler DC, Christensen BC, Marsit CJ, Nelson HH, et al. DNA methylation arrays as surrogate measures of cell mixture distribution. BMC Bioinformatics. 2012;13:1–16.
Hugo HJ, Kokkinos MI, Blick T, Ackland ML, Thompson EW, Newgreen DF. Defining the e-cadherin repressor interactome in epithelial-mesenchymal transition: the PMC42 model as a case study. Cells Tissues Organs. 2011;193:23–40.
Akulenko R, Helms V. DNA co-methylation analysis suggests novel functional associations between gene pairs in breast cancer samples. Hum Mol Genet. 2013;15:3016–22.
Yoder JA, Walsh CP, Bestor TH. Cytosine methylation and the ecology of intragenomic parasites. Trends Genet. 1997;13:335–40.
Wilson AS, Power BE, Molloy PL. DNA hypomethylation and human diseases. Biochim Biophys Acta. 2007;1775:138–62.
Su J, Shao X, Liu H, Liu S, Wu Q, Zhang Y. Genome-wide dynamic changes of DNA methylation of repetitive elements in human embryonic stem cells and fetal fibroblasts. 2012, Genomics. 99(1): 10-7.
Wong NC, Pope BJ, Candiloro I, Korbie D, Trau M, Wong SQ, et al. Supporting data and materials for “Exemplary multiplex bisulfite amplicon data used to demonstrate the utility of Methpat”. GigaScience Database. 2015. http://dx.doi.org/10.5524/100167
We acknowledge Illumina Australia Pty Ltd for a MiSeq Pilot Sequencing Grant. This work was supported, in part, by National Breast Cancer Foundation of Australia (NCBF) grants to AD, EWT, DK and MT (CG-08-07, CG-10-04 and CG-12-07) and NHMRC APP1027527 (EWT, AD, NW, BJWVD). SW was supported by the Melbourne Melanoma Project funded by the Victorian Cancer Agency Translational Research program and established through support of the Victor Smorgon Charitable Fund. Computation time was granted by the Life Sciences Computation Centre (LSCC) at the Victorian Life Sciences Computational Initiative (VLSCI) under grant VR0002. The Murdoch Childrens Research Institute and St. Vincent’s Institute are supported by the Victorian Government Operational and Infrastructure Support Grant.
The authors declare that they have no competing interests.
NCW drafted the manuscript, designed and performed the experiment, and analyzed the data. BJP co-wrote the manuscript, wrote Methpat for visualization and analyzed the data. IC co-wrote the manuscript, designed and performed the experiment. DK co-wrote the manuscript and analyzed the data. MT co-wrote the manuscript. SQW co-wrote the manuscript and performed the experiment. TM co-wrote the manuscript, designed and performed the experiment. BJWVD co-wrote the manuscript and performed the experiment. EWT co-wrote the manuscript. CB co-wrote the manuscript and performed the experiment. SE co-wrote the manuscript and performed the experiment. SRD co-wrote the manuscript and performed the experiment. AD co-wrote the experiment, designed the experiment and analyzed and interpreted the results. All authors read and approved the final manuscript.