0
0
Entering edit mode
@areyoujokingme-9334
Last seen 6.5 years ago
United States

I cannot install RnBeads and ggbio on R version 3.2.2.

ggbio: I get the following error message when I try to install ggbio:

> source("https://bioconductor.org/biocLite.R")
Bioconductor version 3.0 (BiocInstaller 1.16.5), ?biocLite for help
BiocInstaller version 3.0 is too old for R version 3.2.2; remove.packages("BiocInstaller") then
source("http://bioconductor.org/biocLite.R")
A new version of Bioconductor is available after installing the most recent version of R; see
http://bioconductor.org/install
> biocLite("ggbio")
BioC_mirror: http://bioconductor.org
Using Bioconductor version 3.0 (BiocInstaller 1.16.5), R version 3.2.2.
Installing package(s) 'ggbio'
trying URL 'http://bioconductor.org/packages/3.0/bioc/src/contrib/ggbio_1.14.0.tar.gz'
Content type 'application/x-gzip' length 1521091 bytes (1.5 MB)
==================================================

* installing *source* package ‘ggbio’ ...
** R
** inst
Error : object ‘latticeGrob’ is not exported by 'namespace:gridExtra'
* removing ‘/home/lillian/R/x86_64-pc-linux-gnu-library/3.2/ggbio’

Warning message:
In install.packages(pkgs = pkgs, lib = lib, repos = repos, ...) :
installation of package ‘ggbio’ had non-zero exit status

When I try to install RnBeads, I get the following error message:

> source("https://bioconductor.org/biocLite.R")
Bioconductor version 3.0 (BiocInstaller 1.16.5), ?biocLite for help
BiocInstaller version 3.0 is too old for R version 3.2.2; remove.packages("BiocInstaller") then
source("http://bioconductor.org/biocLite.R")
A new version of Bioconductor is available after installing the most recent version of R; see
http://bioconductor.org/install
BioC_mirror: http://bioconductor.org
Using Bioconductor version 3.0 (BiocInstaller 1.16.5), R version 3.2.2.

Warning message:
package ‘RnBeads’ is not available (for R version 3.2.2) 

Can anyone help? Here is my sessionInfo()

> sessionInfo()
R version 3.2.2 (2015-08-14)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 14.04.3 LTS

locale:
[1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C               LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8
[5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8    LC_PAPER=en_US.UTF-8       LC_NAME=C

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base

other attached packages:
[1] gridExtra_2.0.0      BiocInstaller_1.16.5 RColorBrewer_1.1-2   gplots_2.17.0        xtable_1.8-0

loaded via a namespace (and not attached):
[1] tools_3.2.2        gtable_0.1.2       KernSmooth_2.23-15 gdata_2.17.0       grid_3.2.2         caTools_1.17.1
[7] bitops_1.0-6 
ggbio rnbeads rversion3.2.2 • 1.5k views
0
Entering edit mode

bioclite is trying to tell you what to do, you should do it:

BiocInstaller version 3.0 is too old for R version 3.2.2; remove.packages("BiocInstaller") then
source("http://bioconductor.org/biocLite.R")
A new version of Bioconductor is available after installing the most recent version of R; see
http://bioconductor.org/install
0
Entering edit mode

I've tried this many times, and I did it again and nothing changes. I have local R library path at ~/R/x86_64-pc-linuxgnu-library/3.2 and an all users R library path(s) at /usr/local/R/site-library and /usr/lib/R/library/ paths. When I try to remove BiocInstaller, it tries to remove the /usr/local/R/site-library BiocInstaller, and this path is not writable, and so the remove fails. I don't know how to get around this. This shouldn't be hard, but it is proving to be almost impossible for me, haha.

1
Entering edit mode

Do you have root access on your machine? If so, run

sudo R

And then remove BiocInstaller. Then run R as yourself again for the subsequent steps.

0
Entering edit mode

Thanks for all of your input. I ended up deleting R entirely and all libraries in /usr/ and installed a fresh install in a local directory ~/R from source. I think the lesson here is that my initial install of R should have been run with *sudo* apt-get. This, I *think*, would have removed all drama with these kinds of issues. I *think*.

0
Entering edit mode

I would be surprised if "apt-get install" did not require you to run it as root.

More likely is that R itself was run as root at some point and Bioconductor was installed.

1
Entering edit mode

Ideally you wouldn't have the BiocInstaller package installed in any of your system-wide library dirs, over which you apparently have no control. If you don't have sudo, you might ask your sysadmin to remove it for you (and any other Bioconductor packages). But maybe your sysadmin installed all that stuff on purpose, and wants to keep control, in which case you should ask that they be updated.

There are two other alternatives I can think of. First, since you do have a personal library dir, you could download the correct BiocInstaller package by hand and install it in your personal dir:

download.file("http://bioconductor.org/packages/release/bioc/src/contrib/BiocInstaller_1.20.1.tar.gz","BiocInstaller_1.20.1.tar.gz")
install.packages("BiocInstaller_1.20.1.tar.gz", repos = NULL, lib = .libPaths()[1])
## here I am assuming you have your .libPaths() set correctly so the first one is your local dir

By putting the correct version in the 'first' library dir, your R won't see the outdated version. However, this is sort of a bootleg way to do things, as you will still have a fragile install, where a remove.packages("BiocInstaller") will set you back to square one. And if you have other Bioconductor packages in any of the system-wide directories, you won't be able to update them either.

An alternative is to have your own personal install of R/BioC in your home directory, or somewhere else that you have the space and write permissions. In that situation you would only have one library dir, over which you have full control. The downside of that approach is that you have to install and maintain your own R/BioC installation, which depending on the person is either a good idea or a horrible mistake. YMMV.