[BioC] Installing Bioconductor on Linux...

Steffen Möller steffen_moeller at gmx.de
Mon Oct 4 14:46:10 CEST 2010


On 10/04/2010 01:59 PM, James Carman wrote:
>> (2) 'Third party' dependencies need to be satisfied in a rational way,
>> remembering (a) that R packages have complex dependencies with other R
>> packages, and (b) R compiles C (and other) source code and so requires
>> header files associated with third party libraries. Combining these, one
>> can imagine biocLite("biomaRt") failing because XML fails because RCurl
>> fails because the *devel* libcurl (devel required for the curl headers)
>> is not installed. This will be indicated in the output of biocLite, but
>> will require patient inspection of the output to see this. Part of the
>> third party installation process may mean evaluating standard Linux
>> commands (e.g., /sbin/ldconfig), setting environment variables
>> (LD_LIBRARY_PATH ?), and under worst-case scenarios (Rmpi comes to mind)
>> inspecting the R package configure.in / configure.ac script (by
>> downloading the source package from CRAN or Bioconductor) to understand
>> what the requirements are and how they are supposed to be satisfied.
> This is my main concern, the libraries that Bioconductor needs to be
> installed in the operating system.  So, there's no one-shot method of
> getting this done, eh?  I basically need to see what it complains
> about and then figure out how to satisfy the dependency?
the Debian/Ubuntu community is very aware of BioConductor.
And there were several efforts, both manual and automated,
to provide Debian/Ubuntu packages for it ... including all its
C and other external libraries.

The problem is that yet nobody came up with the necessary
amount of spare time to maintain the BioConductor packages.
And we always/usually want to use the latest versions which
then are not too difficult to install manually. The advantage of
Debian is the very clear package management, i.e. you get
things out of the system and have some confidence that the
installation on another system is truly the same.

There is probably a consensus that the package should not
become part of the regular Debian/Ubuntu distribution. What
instead one could try, coming to mind while I write those lines,
is a Debian/Ubuntu package that drags in the binaries
dependencies that are already in Debian/Ubuntu and then
executes the getBioC script in some suitable manner as a
post-inst script. And one could have several such packages
for the various arguments that getBioC could take. Would
you like that? Would I like that? Sounds somewhat error
prone and very much against how R packages today appear
in Debian/Ubuntu and one basically loses the advantage of
getting it all out again. Other ideas?



More information about the Bioconductor mailing list