[Bioc-devel] experiment package submission

[Martin Morgan]

Hi Robert --

On 07/21/2010 11:57 PM, Robert Castelo wrote:
> Martin, a question regarding this issue just when through my mind,
> 
> On Wed, 2010-07-21 at 19:20 -0700, Martin Morgan wrote:
> [...]
>> An experiment data package is additional work, but in exchange you get
>> reproducible, reliable, and documented research.
> 
> i haven't found any especific reference to experiment data package
> submission in
> 
> http://wiki.fhcrc.org/bioc/HowTo/Package_Contribution
> 
> and links thereafter.
> 
> 
> should an experiment data package be submitted following the same
> procedure as a software package? does it go also through a peer-review
> process? are there requirements or guidelines specific for experiment
> data packages?

Often experiment data packages are produced much like in Peter's case --
to support analysis in a particular package or group of packages. They
(the analysis and data packages) are then subject jointly to the preview
process.

Sometimes experiment data packages emerge after the fact, when it
becomes apparent that work flows or packages would benefit from a common
data set. These generally get introduced ad hoc, without formal preview,
but of course the packages are being built and passing R CMD check.

We don't usually start with a 'pure' experiment data package as a
submission -- Bioconductor is not a data repository in that sense -- but
if one were to come in we'd treat it as a new package and submit it to
the same formal acceptance procedure.

Hope that helps,

Martin

> 
> thanks!
> robert.
> 


-- 
Martin Morgan
Computational Biology / Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N.
PO Box 19024 Seattle, WA 98109

Location: Arnold Building M1 B861
Phone: (206) 667-2793