Installation of add-on packages downloaded from this site

This is a short description of the installation of the packages you can download from the main site. For more details, please take a look at the manuals provided by the R project.

Prerequisites

The WGCNA package requires the following packages to be installed: stats, grDevices, utils, matrixStats (0.8.1 or higher), Hmisc, splines, foreach, doParallel, fastcluster, dynamicTreeCut, survival, parallel, preprocessCore, GO.db, impute, and AnnotationDbi. If your system does not have them installed, the easiest way to install them is to issue the following command at the R prompt:

install.packages(c("matrixStats", "Hmisc", "splines", "foreach", "doParallel", "fastcluster", "dynamicTreeCut", "survival", "BiocManager")
BiocManager::install(c("GO.db", "preprocessCore", "impute"));

If you have an old version of R for which BiocManager is not available, you can replace the BiocManager::install call by

source("http://bioconductor.org/biocLite.R")
biocLite(c("GO.db", "preprocessCore", "impute"))

If you plan on using annotation capabilities (such as GOenrichmentAnalysis), we also recommend installing Bioconductor annotation packages. This can be accomplished by the following R commands:

orgCodes = c("Hs", "Mm", "Rn", "Pf", "Sc", "Dm", "Bt", "Ce", "Cf", "Dr", "Gg");
orgExtensions = c(rep(".eg", 4), ".sgd", rep(".eg", 6));
packageNames = paste("org.", orgCodes, orgExtensions, ".db", sep="");

BiocManager::install(c("GO.db", "KEGG.db", "topGO", packageNames, "hgu133a.db", "hgu95av2.db", "annotate", "hgu133plus2.db", "SNPlocs.Hsapiens.dbSNP.20100427", "minet", "OrderedList"))

Replace the call to BiocManager::install by biocLite if the former is not available. If you run an older version of R, the above may not install the newest version of the dynamicTreeCut package. Should you encounter this problem, please manually download and install dynamicTreeCut from this web page.

Installation of the WGCNA package

Installation of the package is simple. Download the appropriate package file from the download page and save it in a directory of your choice. Run an R session as a user with sufficient privileges to install software (see below if you don't have the requisite privileges). At the prompt, type (replace path/to/file with the directory and filename of the downloaded package):

install.packages("path/to/file", repos = NULL, lib=.Library)

For Mac users: if you are compiling and installing a source package, add the argument

type = "source"

to the install.packages command. The additional argument should not be supplied if you are installing the pre-compiled binary bundle (.tgz file). The packages should be automatically unpacked and installed. If you have trouble installing the package, make sure you have installed Apple Xcode Tools, including Xcode command line tools, available from http://developer.apple.com, installed. See the R documentation on http://cran.r-project.org, in particular this section for more details on installation of source packages on OS X.

If you do not have sufficient privileges, you can either (1) ask your system administrator to install the packages for you, or (2) create your own personal R library. You may also wish to create a separate R library to hold oder versions of these and/or other packages. To do that, create a directory on your system where you want the library to reside. Then run an R session, and at the prompt type (replace path/to/file with the directory and filename of the downloaded package and path/to/library with your library directory)

install.packages("path/to/file", repos = NULL, lib = "path/to/library")

Note that if you chose to go this route, you will need to specify your library location every time you load the package, by using

library(package, lib.loc = "path/to/library")

instead of the usual

library(package)

In particular, you will need to modify in this way the example scripts provided at the main page.

Troubleshooting

When downloading and saving the package file, make sure the file is saved without unpacking it. For example, Windows may automatically unzip the file WGCNA_0.75.zip, thus preventing R from installing it correctly. Similarly, Mac OS may automaticall unzip the source file WGCNA_0.75.tar.gz and turn it into WGCNA_0.75.tar, thus preventing R from being able to unpack it. Again, the downloaded package must be saved as is, without any changes to the file.
When installing the package on Mac OS, the following error message may appear at the end of package installation:

cp: /Library/Frameworks/R.framework/Resources/library/R.css: fchmod failed: Operation not permitted

As far as we can tell, this is a harmless error and the package will function normally.

Back to main page

Peter Langfelder, January 2015.