Message Image  

Installing R and Bioconductor Tips (updated with Rgraphviz info)

By Rodney Sparapani posted 04-05-2013 16:00

R is GNU GPL software from the R Project

The Bioconductor Project produces R packages for bioinformatics

Bioconductor versions are tied to specific R versions.  The corresponding Bioconductor version that goes with your R version can be determined by typing the following at the R prompt:
> tools:::.BioC_version_associated_with_R_version
[1] '2.11'
For example, Bioconductor 2.11 is built for recent R versions 2.15.2 and 2.15.3.

Binary distributions of R are easier to install, but they are not optimized for the specific CPU that you are running.  Binary distributions are available for Mac OS X, Windows and many flavors of Linux.  However, on Linux (like UNIX), I prefer the source distribution so that you can optimize for the specific CPU that you are running.

You can get R, whether binary or source, from your favorite
Comprehensive R Archive Network (CRAN) mirror like StatLib

There are groupings of R packages by topic such as
Statistical Genetics
Here's how to install R packages from the StatLib CRAN mirror via the R prompt.  To install the R package named genetics (and R package dependencies, if any):
> options(repos=c(CRAN=""))
> install.packages("genetics", dependencies=TRUE)

To install all R packages (this will take a while):
> install.packages(available.packages()[,1])

To search the R help and installed R package help:
> help.start()

Bioconductor packages are R packages so some of the same remarks apply.  However, Bioconductor has their own simple installation procedure.  To install the package named limma (and R or Bioconductor package dependencies, if any):

> source("")
> biocLite("limma")

To install all Bioconductor packages (takes a while):
> biocLite(all_group())

Some Bioconductor packages require the R package
Rgraphviz which depends on graphviz.  graphviz is a free
graph visualization software for Windows, Mac OS X and
UNIX/Linux originally developed by AT&T

It is much easier to install Rgraphviz/graphviz on Windows
and Mac OS X since the binary distribution of Rgraphviz
"bundles" graphviz.

To install Rgraphviz/graphiz on Windows and Mac OS X:
> biocLite("Rgraphviz")


Related Resources

No Related Resource entered.