Methylation plotter

A dynamic Web tool for easy methylation data visualization



Methylation plotter allows to generate publication-ready, high quality graphics summarizing methylation data in lollipop- and heatmap-like plots. Some descriptive stats and comparisons are provided.

Please cite our paper Methylation plotter: a web tool for dynamic visualization of DNA methylation data Source Code for Biology and Medicine 2014, 9:11.

Further information, including example datasets, is available at the documentation webpage.

x

Methylation plotter address update

The preferred URL to access Methylation plotter has changed.

Please update your bookmark http://maplab.cat/methylation_plotter. Thanks!

Methylation data

Methylation file must be a tab separated file with the following structure:
  • Methylation levels must be indicated in percentage (from 0 to 1).
  • Each row contains the methylation levels for a sample. The last row must indicate the position of the CpGs (positions can be unsorted).
  • First column indicates the sample identifier. Each middle column contains the methylation levels for a CpG.
  • Last column indicates at which group (i.e., control, treatment) do the sample belongs.
Mind that the maximum number of both CpGs and samples to be plotted is 100. A toy dataset is available here.

Interval file is an optional, tab separated file indicating the intervals of methylation for the gray scale color legend. A toy dataset is available here . If not provided, a predefined set of ten intervals ranging from black (fully methylated CpGs) to white (fully unmethylated CpGs) corresponds to unmethylated CpGs will be used.

Plotting options

There are two major plot types:
  • circles offers a lollipop-style plot. Each CpG is represented by a circle filled with a shade of gray according to its methylation status. Missing data are represented by crosses. By default the lollipop-syle plot takes into account the distance between CpGs (the longer the distance between them, the longer space between lollipops); this behaviour is included as the proportional circles option. The nonproportional circles disables this behaviour, just spacing the CpGs regardless of the actual position.
  • grid represents methylation data in a heatmap-like manner. Missing data are represented in blue. This does not take into account the distance between CpGs that's why it is more useful when plotting a large number of CpGs.


There are four ways to sort the data:
  • as-is plots data in the same order as it was in the input file.
  • by methylation level sorts data by the mean of the samples from more methylated to more unmethylated.
  • by group sorts data using the group column of the input file alphabetically.
  • unsupervised clustering sorts data implementing an unsupervised clustering analysis.

Methylation plot size

The height and width sliders allow the fine tuning of the final plot size.
Methylation plot

Save image as PNG Save image as PDF
Methylation profile plot
Each line represents for each group of samples the methylation mean for each position. Asterisks indicate a statistical significance as calculated by the Kruskal-Wallis test.
Save image as PNG Save image as PDF
Dendrogram
A dendrogram showing the unsupervised clustering of the data. The groups provided by the user are highlighted with different colors.
Save image as PNG Save image as PDF
Boxplot
Boxplot for each group of samples.
Save image as PNG Save image as PDF
Summary by sample
Summary table with the mean, the standard deviation, the minimum, the maximum and the number of missing data for each sample.

Save data
Descriptive analysis
Summary table with the mean, the standard deviation, the minimum, the maximum and the number of missing data for each position and group of samples. If the samples are categorized in two or more groups, it also shows the p value of the Kruskal-Wallis test. This test is the non-parametric version of the ANOVA (one-way analysis of variance) and tests whether samples originate from the same distribution. If the test is statistically significant (p value less than 0.05) it means that at least one of the samples is different from the other samples.

Save data

© IMPPC, 2014 - Designed and maintained by Izaskun Mallona and Anna Díez-Villanueva. Logo by Julien Douet.