The purpose of MethylExpress is to allow for visual investigation of physical characteristics of methylation of DNA. Examples of this are counting the amount of CpG islands that are present in a DNA strand and highlighting where they are in the DNA strand, highlighting where methylation occurs in a DNA strand, and graphically viewing the differences in gene expression for two DNA strands with different nucleotides being methylated.
This R package is unique from other R packages that deal with methylation data as it allows for visual understanding and interpretation of methylation data. While there are packages that, for example, allow for the amount of CpG islands to be calculated for a given DNA strand, there are no packages that allow the user to visually see where in the DNA strand the CpG islands occur. Not only does this R package allow for visualization, but it combines the ideas of other package into a single package, while also adding more functionality!
To download the package:
require("devtools")
devtools::install_github("diannamcallister/MethylExpress", build_vignettes = TRUE)
library("MethylExpress")To run the shinyApp:
MethylExpress::runMethylExpress()ls("package:MethylExpress")
data(package="MethylExpress")There are 4 datasets that are able to be accessed by the user:
1. PossibleCpGIslands is a fabricated (fake) dataset, that is a DNA strand in the form of a string of nucleotides, that can be used as an example dataset to calculate the amount of CpG islands are present in a DNA strand.
2. MethylationObservation is a dataset retrieved from a study performed on Ceratina calcarata at the University of Georgia, where mapping of methylation and quantification of gene expression was completed (GSE111611). The DNA sequence before and after bisulfite conversion are found in this dataset.
3. BeforeBariatricSurgery is a dataset retrieved from a study performed at INSERM on overweight adults who would be recieving bariatric surgery (GSE66306). This dataset includes the patients RNAseq data before the surgery.
4. AfterBariatricSurgery is a dataset retrieved from a study performed at INSERM on overweight adults who had recieved bariatric surgery. (GSE66306).This dataset includes the patients RNAseq data after the surgery.
There are 5 functions that are able to be accessed by the user:
1. differencesInGeneExpression allows for users to be able to compare two different RNAseq data from two different DNA strands (preferably with different methylation) and see the n genes with the largest difference in expression between the two RNAseq datasets in a graph. The user is able to specify n, which is the number of genes shown in the graphical output.
2. findCpGIslands allows for users to be able to determine how many CpG islands are present in their DNA strand, as well as view where the CpG islands are in the DNA strand as this function will output the nucleotides of the given strand with the CpG islands highlighted.
3. findDifferencesInDNASequence allows for users to be able to determine which nucleotides between two strands of DNA are different. This function will output the two DNA strands with the nucleotides that differ between the strands highlighted.
4. findMethylatedCytosines allows for user to compare a DNA sequence before and after bisulfite conversion, and determine which cytosine nucleotides are methylated or not. This function will output the original DNA strand (before bisulfite conversion) with the cytosines that are methylated highlighted.
5. runMethylExpress allows for the user to run the shiny app for MethylExpress so that they can interact with all of the functionalities of the package through a GUI instead of the R commandline.
For more details about the functions, please take a look at the vignette for this package:
browseVignettes("MethylExpress")The package tree structure is provided below:
- MethylExpress
|- MethylExpress.Rproj
|- DESCRIPTION
|- NAMESPACE
|- LICENSE
|- README
|- data
|- AfterBariatricSurgery.rda
|- BeforeBariatricSurgery.rda
|- MethylationObservation.rda
|- PossibleCpGIslands.rda
|- inst
|- extdata
|- MethylExpressOverview.png
|- shiny-scripts
|- app.R
|- man
|- AfterBariatricSurgery.Rd
|- BeforeBariatricSurgery.Rd
|- countCpGIslands.Rd
|- differencesInGeneExpression.Rd
|- DNASequenceHighlights.Rd
|- findCpGIslands.Rd
|- findDifferenceOfExpression.Rd
|- findDifferencesInDNASequence.Rd
|- findMatchingSections.Rd
|- findMethylatedCytosines.Rd
|- highlight.Rd
|- MethylationObservation.Rd
|- observedAndExpected.Rd
|- PossibleCpGIslands.Rd
|- runMethylExpress.Rd
|- R
|- data.R
|- differencesInGeneExpression.R
|- findCpGIslands.R
|- findDifferencesInDNASequence.R
|- findMethylatedCytosines.R
|- runMethylExpress.R
|- vignettes
|- differencesInGeneExpressionExResult.png
|- findCpGIslandsExResult.png
|- findDifferencesInDNASequenceExResult.png
|- findMethylatedCytosinesExResult.png
|- MethylExpressVignette.Rmd
|- runMethylatedExpressExResult.png
|- tests
|- testthat.R
|- testthat
|- test-DifferencesInGeneExpression.R
|- test-FindCpGIslands.R
|- test-FindDifferencesInDNASequence.R
|- test-FindMethylatedCytosines.RAn overview of the package is illustrated below:
The author of this package is Dianna McAllister.
For the function DifferencesInGeneExpression, I use the R package ggplot2 to be able to plot the different amounts of gene expression.
For the functions FindCpGIslands, FindDifferencesInDNASequence, and FindMethylatedCytosines, I use rstudioapi to be able to display my markdown files (with the highlights) in the Rstudio Viewer panel.
H. Wickham. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York, 2016.
Hunt, B. G. (2018). GEO Accession viewer. Retrieved November 17, 2020, from https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE111611
Kevin Ushey, JJ Allaire, Hadley Wickham and Gary Ritchie (2020). rstudioapi: Safely Access the RStudio API. R package version 0.13. https://CRAN.R-project.org/package=rstudioapi
Poitou C, Perret C, Mathieu F, Truong V et al. Bariatric Surgery Induces Disruption in Inflammatory Signaling Pathways Mediated by Immune Cells in Adipose Tissue: A RNA-Seq Study. PLoS One 2015;10(5):e0125718. PMID: 25938420
R Core Team (2019). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/.
This package was developed as part of an assessment for 2020 BCB410H: Applied Bioinformatics, University of Toronto, Toronto, CANADA.