[DAS] More annotation/Reference servers please

[Doug Hyatt]

"Cheung, Foo" wrote:

> Hi all
> sorry to say the obvious...
>
> What annotations do people what?, what client functionality do scientist
> want?, what is going to get scientists to adopt this technology? Why should
> they adopt DAS? The bottom line is, DAS has to provide something that
> "forces" scientists to use it, something that goin to break their habits of
> going from one annotation web site to another. Its time to start thinking
> Microsoft style :-)
> I am not sure what that is?
>

The analogy I have made with DAS around here is that it's like a bunch of us
are
working on the specifications of HTML as well as making a bunch of
cool HTML web pages.... but until we have a Mosaic/Netscape, the specification
remains exactly that: a specification.  I'd say we need "Dreamweaver"-like
tools
for DAS also that really make it easy to build a server, but this is hard...
people
have their data in all sorts of different formats/databases.

The principal problem I see is that bioinformaticists/biologists typically get
their
data in one of 2 ways (and they LIKE doing it these ways too):
  (1) downloading in bulk from some FTP site (i.e. going to Ensembl and
grabbing
  the fasta file of all their transcripts) so you can mess around with it
yourself
  (2) querying/blasting against a particular region of interest, i.e. give me a
little
  box I can type words into and get back results or let me go to a particular
  gene I'm interested in and study it, etc etc.

DAS doesn't really fit into these 2 models, since it isn't really queryable
except "geographically", i.e. gimme annotations in this region.  As has been
repeatedly stressed, though, queryability was not DAS's function.  It was just
designed to make it easy to display annotations from different sources, which
it does very well.

In general, I think it's hard to get people to 'browse' geographically.  They
usually either
just want your stuff and come grab it in bulk.... or they have a particular
something they're looking for and they ask about it through a search box or
a BLAST.  What would be cool, imo, is a query engine on top of a DAS
framework, where you could ask about a particular protein and have a
client just pop up all the Ensembl, Genome Channel, NCBI, etc. etc. versions
of that gene in DAS.

> I believe that DAS is way cool, there are some amazingly bright software
> developers working on the java clients and some top class scientists pushing
> DAS but I would like someone to address how to get "customers" lining up
> each morning at 8:50am waiting to get in the front door?
>

It's hard... it requires getting people to think in new ways.  Right now
a good many people still want to just ftp data and reformat it themselves if
they're really
going to use it.... and if they're not going to use the bulk data, they want to

query something specific that they already have a clue about.

>
> I think this has been addressed previously but was not really resolved.
>

regards,
doug


Doug Hyatt
Computational Biology
Oak Ridge National Laboratory
hyattpd@ornl.gov

>
> Foo
>
> _______________________________________________
> DAS mailing list
> DAS@biodas.org
> http://biodas.org/mailman/listinfo/das