Hi, Richard,

I see you are looking in the main toolshed. I cant speak for fcarmia's one but it doesn't do any dependency installation. The 'digital_dge' one was deprecated - I've now deleted it.

The differential count models package in the statistics section of the test toolshed - http://testtoolshed.g2.bx.psu.edu/view/fubar/differential_count_models is ready for testing so if you have time, please try it. It installs all dependencies including R and bioconductor automatically. Installation through the galaxy admin interface should work automatically but be warned - it takes a very long time to install all the dependencies including compiling R and installing Bioconductor. You will need to generate count matrices from bam files - there's an htseq_bams_to_count_matrix tool to do that which also installs all dependencies - also in the statistics section.

On Thu, Aug 15, 2013 at 1:17 AM, Richard Kuo <izenkuo@gmail.com> wrote:

Hello,
I have some questions about the edgeR (fcaramia) and digital_dge (fubar) repositories in tool shed.

What is the difference between the two respositories in terms of being able to run edgeR?

What is the proper way to install R so that the tool wrappers can use edgeR?

What versions of R and edgeR do each of the repositories work with? Similaraly why is it that the repositories do not have information about which versions of R and edgeR are compatible?

I have tried both repositories and have installed R and loaded the edgeR library but the tools do not seem to work. There are no errors coming up but the outputs are empty. The input file works with edgeR ran the traditional way through R (without Galaxy).

Thank you,
Richard

____________________________________________