Generate a HTML/PDF report exploring edgeR results

Description

This function generates a HTML report with exploratory data analysis plots for edgeR results created. Other output formats are possible such as PDF reports but they lose the interactivity. Users can easily append to the report by providing a R Markdown file to customCode, or can customize the entire template by providing an R Markdown file to template.

Usage

1
2
3
4
5
6
7
edgeReport(dge, object, project = "", intgroup, colors = NULL,
  pAdjustMethod = "BH", alpha = 0.1, independentFiltering = FALSE, filter,
  theta, filterFun, nBest = 500, nBestFeatures = 20, customCode = NULL,
  outdir = "edgeRexploration", output = "edgeRexploration",
  browse = interactive(), device = "png", template = NULL,
  searchURL = "http://www.ncbi.nlm.nih.gov/gene/?term=", theme = NULL,
  digits = 2, ...)

Arguments

dge

A DGEList object.

object

A DGEExact or DGELRT object that contains p-values stored in object$table$PValue.

project

The title of the project.

intgroup

interesting groups: a character vector of names in colData(x) to use for grouping. This parameter is passed to functions such as plotPCA.

colors

vector of colors used in heatmap. If NULL, then a a default set of colors will be used. This argument is passed to pheatmap.

pAdjustMethod

the method to use for adjusting p-values, see p.adjust. This argument will be passed to results.

alpha

the significance cutoff used for optimizing the independent filtering (by default 0.1). If the adjusted p-value cutoff (FDR) will be a value other than 0.1, alpha should be set to that value. This argument will be passed to results.

independentFiltering

logical, whether independent filtering should be applied automatically. By default it's set to FALSE in contrast with the default used in results to match edgeR's behavior.

filter

the vector of filter statistics over which the independent filtering will be optimized. By default the logCPM will be used if independentFiltering is set to TRUE. It can also be a length 1 character vector specifying one of the column names of object$table.

theta

the quantiles at which to assess the number of rejections from independent filtering. This argument is passed results.

filterFun

an optional custom function as described in results.

nBest

The number of features to include in the interactive table. Features are ordered by their adjusted p-values.

nBestFeatures

The number of best features to make plots of their counts. We recommend a small number, say 20.

customCode

An absolute path to a child R Markdown file with code to be evaluated before the reproducibility section. Its useful for users who want to customize the report by adding conclusions derived from the data and/or further quality checks and plots.

outdir

The name of output directory.

output

The name of output HTML file (without the html extension).

browse

If TRUE the HTML report is opened in your browser once it's completed.

device

The graphical device used when knitting. See more at http://yihui.name/knitr/options (dev argument).

template

Template file to use for the report. If not provided, will use the default file found in DESeq2Exploration/DESeq2Exploration.Rmd within the package source.

searchURL

A url used for searching the name of the features in the web. By default http://www.ncbi.nlm.nih.gov/gene/?term= is used which is the recommended option when features are genes. It's only used when the output is a HTML file.

theme

A ggplot2 theme to use for the plots made with ggplot2.

digits

The number of digits to round to in the interactive table of the top nBestFeatures. Note that p-values and adjusted p-values won't be rounded.

...

Arguments passed to other methods and/or advanced arguments. Advanced arguments:

software

The name of the package used for performing the differential expression analysis. Either DESeq2 or edgeR.

dge

A DGEList object. NULL by default and only used by edgeReport.

theCall

The function call. NULL by default and only used by edgeReport.

output_format

Either html_document, pdf_document or knitrBootstrap::bootstrap_document unless you modify the YAML template.

clean

Logical, whether to clean the results or not. Passed to render.

Details

Set output_format to 'knitrBootstrap::bootstrap_document' or 'pdf_document' if you want a HTML report styled by knitrBootstrap or a PDF report respectively. If using knitrBootstrap, we recommend the version available only via GitHub at https://github.com/jimhester/knitrBootstrap which has nicer features than the current version available via CRAN.

If you modify the YAML front matter of template, you can use other values for output_format.

This report is similar to the one created by DESeq2Report with two additional plots exclusive for edgeR results. We designed the reports to be very similar intentionally and use the Bioconductor package DEFormats to achieve this goal.

Value

An HTML report with a basic exploration for the given set of edgeR results.

Author(s)

Leonardo Collado-Torres

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
## Create example data using DEFormats
library('DEFormats')
set.seed(20160407)
counts <- simulateRnaSeqData()
group <- rep(c("A", "B"), each = 3)

## Create DGEList object
library('edgeR')
dge <- DGEList(counts, group = group)

## Perform DE analysis with edgeR
design <- model.matrix( ~ group)
dge <- estimateDisp(dge, design)
fit <- glmFit(dge, design)
lrt <- glmLRT(fit, coef = 2)

## The output will be saved in the 'edgeReport-example' directory
dir.create('edgeReport-example', showWarnings = FALSE, recursive = TRUE)

## Generate the HTML report
report <- edgeReport(dge, lrt, project = 'edgeR-example', intgroup = 'group',
    outdir = 'edgeReport-example')

if(interactive()) {
    ## Browse the report
    browseURL(report)
}

## Not run: 
## Note that you can run the example using:
example('edgeReport', 'regionReport', ask=FALSE)

## End(Not run)