[BioC] usage stats for bioconductor

Martin Morgan mtmorgan at fhcrc.org
Fri Dec 17 02:08:48 CET 2010


On 12/16/2010 04:38 PM, Patrick Leyshock wrote:
> I had the pleasure of participating in a great Bioconductor workshop last
> week.  I'm afraid I neglected to ask while up in Seattle, but are there any
> statistics about Bioconductor usage among genomics researchers?  E.g. info
> on "market" penetration, adoption rates over time, etc.  I snooped around
> the web but could not find any.

Hi Patrick --

Glad you enjoyed the course. The annual reports

http://bioconductor.org/about/annual-reports/

provide some indication. Seems like there are several metrics --
literature citation, package downloads, mailing list activity, web site
activity. Each of these are challenging to measure, e.g.,

* Is a citation of the 'limma' package also a citation of Bioconductor?
What about a publication from the limma authors different from the one
mentioned in citation('limma')? The annual report doesn't do literature
citations justice.

* While biomaRt is at the top of the package downloads stats page

http://bioconductor.org/packages/stats/index.html

we know on the one hand from our logs that this is because of an
enthusiastic cluster somewhere in the midwest, and on the other that
we're not tracking downloads through our mirrors or through secondary
distribution.

* Our mailing lists are very active, but we don't know what fraction of
the community this represents; lurking on stack overflow / twitter /
etc. suggests little overlap, implying large population size.

Our updated web site includes google analytics, and while these aren't
publicly available, we do see for the last 30 days:

* about 120,000 page views from 18,000 unique visitors (38k total visits)

* visits from the US (13k), UK (3k), Germany (3k), China (2k), France
(1.5k), Canada (1.2k), Italy (1.1k), Japan (1k), Spain (940), India
(860), and 104 other countries.

We also know what people do on the site, e.g., on the 'Workflows' about
40% of page views are about oligo arrays, vs. 25% sequencing, 15%
annotation. The 'Install' page has the highest bounce rate -- either
people find exactly what they want or are entirely put off.

This doesn't really get at market penetration, since we don't have a
handle on total market size. Informally, one can scholar.google.com for
Bioconductor (11,500), Bioperl (2,200), Biopython (645), "UCSC Genome
Browser" (9,070), ...

Martin

> 
> Thanks, Patrick
> 
> 	[[alternative HTML version deleted]]
> 
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at r-project.org
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor


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

Location: M1-B861
Telephone: 206 667-2793



More information about the Bioconductor mailing list