[BioC] package structure

[Robert M. Flight]

Hi All,

I'm looking for some advice on package structure and dependencies.

I'm working on a package that allows for different types of
meta-analyses between different datasets. As part of the original
development leading up to creating a package, I also generated a
series of functions that do a lot of work to generate navigable
results using HTML imagemaps.

However, the visualizations depend currently on Rgraphviz (layouts)
and Cairo (PNG figure generation). The results can be summarized in
different ways than the visualizations I have developed (potentially
using Cytoscape, or just as a table), and I want to make the package
as easy to install and use as possible. Therefore, if I can make the
analysis package not require Rgraphviz and Cairo, I think that would
be good.

Would it be advisable to create a second package containing the
functions necessary for creating the visualizations? Otherwise I
introduce dependencies that may not even be used by the users of the
package. And I don't know of any other proper way (i.e. besides
"require", which I believe is to be avoided in packages) to call other
packages than having them in the DESCRIPTION and NAMESPACE files, and
I believe anything in NAMESPACE requires that the package be
installed.

I would appreciate others thoughts on this.

Thanks,

-Robert

Robert M. Flight, Ph.D.
University of Louisville Bioinformatics Laboratory
University of Louisville
Louisville, KY

PH 502-852-1809 (HSC)
PH 502-852-0467 (Belknap)
EM robert.flight at louisville.edu
EM rflight79 at gmail.com

Williams and Holland's Law:
       If enough data is collected, anything may be proven by
statistical methods.