[DAS2] Human prototype server up
Helt,Gregg
Gregg_Helt at affymetrix.com
Thu Apr 21 23:04:10 UTC 2005
You need to do an anonymous check-out the genoviz CVS module at
http://sourceforge.net/projects/genoviz/ and build the class files.
This assumes you have a Java SDK installed, version 1.4 or later.
You'll also need the xerces XML parser jar.
The test UI I'm working on for the DAS2 client is in the class
com.affymetrix.igb.view.Das2LoadView. Once you've compiled the source,
you should be able to run it from the command line like this:
[java executable] -classpath "[path for genoviz];[path for xerces]"
com.affymetrix.igb.view.Das2LoadView
On my machine this translates to:
java -classpath
"c:/projects/genoviz/classes;c:/projects/genoviz/lib/xerces.jar"
com.affymetrix.igb.view.Das2LoadView
You're presented with a cascading series of choices: DAS2 servers -->
DAS2 sources --> DAS2 versions --> DAS2 regions, and your choice
triggers the appropriate DAS2 query to populate the next choice in the
cascade. Once a version is selected, a DAS2 types query populates a
checkbox list of DAS2 types. You can also test feature queries with
overlap and type filters by entering the min and max for the overlap
filter, selecting which types to filter by, and clicking on "Load
Features". This will bring up a rudimentary visualization of the
sequence and the feature hierarchies returned by the query. By default
it shows the whole length of the sequence, so you'll probably want to
zoom in using the slider on the top and the scrollbar on the bottom.
This viewer shows feature hierarchies as nested outlined boxes. For
client-side data model, there are two additional levels of hierarchy
added above the root features -- the parent of the root features is a
request feature encompassing all the features returned by a single
feature request, and the grandparent is a container for all feature
requests of a given type on a single sequence (the model underneath this
breaks requests for multiple types into separate request for each type).
I just checked interaction of Das2LoadView with DAS2 test server at
das.biopackages.net, and it looks like there are still some issues with
correct handling of feature child/parent relationships. I'll try and
figure out if it's a problem on the client end.
thanks,
gregg
> -----Original Message-----
> From: Allen Day [mailto:allenday at ucla.edu]
> Sent: Thursday, April 21, 2005 2:31 PM
> To: Helt,Gregg; Erwin, Ed
> Cc: das2 at portal.open-bio.org; Andrew Dalke
> Subject: RE: [DAS2] Human prototype server up
>
> Gregg,
>
> Please post your instructions for how to use your server test
application.
> Thanks.
>
> -Allen
>
> On Thu, 21 Apr 2005, Helt,Gregg wrote:
>
> > Cool, great to see a public server up and running!
> >
> > However, I'm having trouble with feature queries. For example:
> >
http://das.biopackages.net/das/genome/chado/chado-Hsa-17/feature?overlap
> > s=chr21/21500000:22500000;type=0000110
> >
> > Which is an example of the most typical query IGB will be making --
one
> > overlap filter combined with one type filter. But this query
returns
> > with a single <FEATURELIST> line, no <FEATURES>. I've tried it for
> > every type returned by the type query and get the same results.
> >
> > Is there something wrong with my query syntax?
> >
> > If this is not a good region to test, can you send me an example
feature
> > query URL where you know what the response should be?
> >
> > thanks!
> > gregg
> >
> > > -----Original Message-----
> > > From: das2-bounces at portal.open-bio.org
> > [mailto:das2-bounces at portal.open-
> > > bio.org] On Behalf Of Allen Day
> > > Sent: Wednesday, April 20, 2005 10:42 PM
> > > To: das2 at portal.open-bio.org
> > > Subject: [DAS2] Human prototype server up
> > >
> > > Contains UCSC Hg17 knownGene, as well as all Affymetrix "target"
> > sequences
> > > aligned to the genome w/ > 90% identity.
> > >
> > > http://das.biopackages.net/das/genome/chado/chado-Hsa-17/region
> > >
> > > Here's a URL to get you going. This is running on 2 hosts: a dual
> > > opteron/8GB webserver frontend and a dual opteron/16GB database
> > backend.
> > > There are no other applications on these hosts right now, so
hammer
> > away.
> > >
> > > You might want to use HTTP 1.1 keepalive though, the responses can
be
> > > *really* slow for /feature requests on large regions. The reason
for
> > this
> > > is the feature/part subpart requests are done feature by feature
due
> > to
> > > the 1-many relationship... I'm still trying to figure out how to
> > improve
> > > this.
> > >
> > > Components that aren't working in even a basic way:
> > >
> > > * alignments
> > > * assemblies
> > > * individual /feature/<id> requests
> > > * versioned source response w/ supported formats
> > > * most feature filters (overlaps, type, and exacttype work)
> > >
> > > -Allen
> > > _______________________________________________
> > > DAS2 mailing list
> > > DAS2 at portal.open-bio.org
> > > http://portal.open-bio.org/mailman/listinfo/das2
> >
>
> On Thu, 21 Apr 2005, Allen Day wrote:
>
> > This is a good region. I just realized the type ids coming back in
the
> > /type response don't match the type ids given in the /feature
response.
> > I'm in the process of switching over from using table PKs for types
to
> > using type names, so rather than using a filter "type=9770" you can
> > instead us a filter "type=sofa.ontology:CDS" (the "sofa.ontology"
string
> > is the term's namespace. SO has already changed it to "SOFA" in
their
> > most recent SOFA file in CVS, but I haven't updated yet).
> >
> > So, sorry about that. I just change thed /type response back to
give
> > primary keys until I get names working. All the types should now
work
> > with the type= filter. Here's a quick document you can grab, a
> subregion
> > of what you asked for in your request that has something interesting
> > inside:
> >
> > http://das.biopackages.net/das/genome/chado/chado-Hsa-
> 17/feature?overlaps=chr21/21292691:21324284;type=9688
> >
> > -Allen
> >
> >
> >
> > On Thu, 21 Apr 2005, Helt,Gregg wrote:
> >
> > > Cool, great to see a public server up and running!
> > >
> > > However, I'm having trouble with feature queries. For example:
> > > http://das.biopackages.net/das/genome/chado/chado-Hsa-
> 17/feature?overlap
> > > s=chr21/21500000:22500000;type=0000110
> > >
> > > Which is an example of the most typical query IGB will be making
--
> one
> > > overlap filter combined with one type filter. But this query
returns
> > > with a single <FEATURELIST> line, no <FEATURES>. I've tried it
for
> > > every type returned by the type query and get the same results.
> > >
> > > Is there something wrong with my query syntax?
> > >
> > > If this is not a good region to test, can you send me an example
> feature
> > > query URL where you know what the response should be?
> > >
> > > thanks!
> > > gregg
> > >
> > > > -----Original Message-----
> > > > From: das2-bounces at portal.open-bio.org
> > > [mailto:das2-bounces at portal.open-
> > > > bio.org] On Behalf Of Allen Day
> > > > Sent: Wednesday, April 20, 2005 10:42 PM
> > > > To: das2 at portal.open-bio.org
> > > > Subject: [DAS2] Human prototype server up
> > > >
> > > > Contains UCSC Hg17 knownGene, as well as all Affymetrix "target"
> > > sequences
> > > > aligned to the genome w/ > 90% identity.
> > > >
> > > > http://das.biopackages.net/das/genome/chado/chado-Hsa-17/region
> > > >
> > > > Here's a URL to get you going. This is running on 2 hosts: a
dual
> > > > opteron/8GB webserver frontend and a dual opteron/16GB database
> > > backend.
> > > > There are no other applications on these hosts right now, so
hammer
> > > away.
> > > >
> > > > You might want to use HTTP 1.1 keepalive though, the responses
can
> be
> > > > *really* slow for /feature requests on large regions. The
reason
> for
> > > this
> > > > is the feature/part subpart requests are done feature by feature
due
> > > to
> > > > the 1-many relationship... I'm still trying to figure out how to
> > > improve
> > > > this.
> > > >
> > > > Components that aren't working in even a basic way:
> > > >
> > > > * alignments
> > > > * assemblies
> > > > * individual /feature/<id> requests
> > > > * versioned source response w/ supported formats
> > > > * most feature filters (overlaps, type, and exacttype work)
> > > >
> > > > -Allen
> > > > _______________________________________________
> > > > DAS2 mailing list
> > > > DAS2 at portal.open-bio.org
> > > > http://portal.open-bio.org/mailman/listinfo/das2
> > >
> > _______________________________________________
> > DAS2 mailing list
> > DAS2 at portal.open-bio.org
> > http://portal.open-bio.org/mailman/listinfo/das2
> >
>
> On Thu, 21 Apr 2005, Ed Erwin wrote:
>
> > That response comes back with some "%20" things in it. That seems
like
> > a bug to me, but is it?
> >
> > <PROP ptype="property/protein%20binding"/>
> >
> >
> > Allen Day wrote:
> > > This is a good region. I just realized the type ids coming back
in
> the
> > > /type response don't match the type ids given in the /feature
response.
> > > I'm in the process of switching over from using table PKs for
types to
> > > using type names, so rather than using a filter "type=9770" you
can
> > > instead us a filter "type=sofa.ontology:CDS" (the "sofa.ontology"
> string
> > > is the term's namespace. SO has already changed it to "SOFA" in
their
> > > most recent SOFA file in CVS, but I haven't updated yet).
> > >
> > > So, sorry about that. I just change thed /type response back to
give
> > > primary keys until I get names working. All the types should now
work
> > > with the type= filter. Here's a quick document you can grab, a
> subregion
> > > of what you asked for in your request that has something
interesting
> > > inside:
> > >
> > > http://das.biopackages.net/das/genome/chado/chado-Hsa-
> 17/feature?overlaps=chr21/21292691:21324284;type=9688
> > >
> > > -Allen
> > >
> > >
> > >
> > > On Thu, 21 Apr 2005, Helt,Gregg wrote:
> > >
> > >
> > >>Cool, great to see a public server up and running!
> > >>
> > >>However, I'm having trouble with feature queries. For example:
> > >>http://das.biopackages.net/das/genome/chado/chado-Hsa-
> 17/feature?overlap
> > >>s=chr21/21500000:22500000;type=0000110
> > >>
> > >>Which is an example of the most typical query IGB will be making
--
> one
> > >>overlap filter combined with one type filter. But this query
returns
> > >>with a single <FEATURELIST> line, no <FEATURES>. I've tried it
for
> > >>every type returned by the type query and get the same results.
> > >>
> > >>Is there something wrong with my query syntax?
> > >>
> > >>If this is not a good region to test, can you send me an example
> feature
> > >>query URL where you know what the response should be?
> > >>
> > >> thanks!
> > >> gregg
> > >>
> > >>
> > >>>-----Original Message-----
> > >>>From: das2-bounces at portal.open-bio.org
> > >>
> > >>[mailto:das2-bounces at portal.open-
> > >>
> > >>>bio.org] On Behalf Of Allen Day
> > >>>Sent: Wednesday, April 20, 2005 10:42 PM
> > >>>To: das2 at portal.open-bio.org
> > >>>Subject: [DAS2] Human prototype server up
> > >>>
> > >>>Contains UCSC Hg17 knownGene, as well as all Affymetrix "target"
> > >>
> > >>sequences
> > >>
> > >>>aligned to the genome w/ > 90% identity.
> > >>>
> > >>>http://das.biopackages.net/das/genome/chado/chado-Hsa-17/region
> > >>>
> > >>>Here's a URL to get you going. This is running on 2 hosts: a
dual
> > >>>opteron/8GB webserver frontend and a dual opteron/16GB database
> > >>
> > >>backend.
> > >>
> > >>>There are no other applications on these hosts right now, so
hammer
> > >>
> > >>away.
> > >>
> > >>>You might want to use HTTP 1.1 keepalive though, the responses
can be
> > >>>*really* slow for /feature requests on large regions. The reason
for
> > >>
> > >>this
> > >>
> > >>>is the feature/part subpart requests are done feature by feature
due
> > >>
> > >>to
> > >>
> > >>>the 1-many relationship... I'm still trying to figure out how to
> > >>
> > >>improve
> > >>
> > >>>this.
> > >>>
> > >>>Components that aren't working in even a basic way:
> > >>>
> > >>> * alignments
> > >>> * assemblies
> > >>> * individual /feature/<id> requests
> > >>> * versioned source response w/ supported formats
> > >>> * most feature filters (overlaps, type, and exacttype work)
> > >>>
> > >>>-Allen
> > >>>_______________________________________________
> > >>>DAS2 mailing list
> > >>>DAS2 at portal.open-bio.org
> > >>>http://portal.open-bio.org/mailman/listinfo/das2
> > >>
> > > _______________________________________________
> > > DAS2 mailing list
> > > DAS2 at portal.open-bio.org
> > > http://portal.open-bio.org/mailman/listinfo/das2
> >
>
> On Thu, 21 Apr 2005, Ed Erwin wrote:
>
> >
> > This is a question for Andrew Dalke.
> >
> >
> > Allen Day wrote:
> > > I don't know, is it? I have found the description of properties
and
> the
> > > /property response document examples to be extremely unclear. I
opted
> to
> > > URI escape here, but maybe it's not the right thing to do.
> > >
> > > -Allen
> > >
> > >
> > > On Thu, 21 Apr 2005, Ed Erwin wrote:
> > >
> > >
> > >>That response comes back with some "%20" things in it. That seems
> like
> > >>a bug to me, but is it?
> > >>
> > >> <PROP ptype="property/protein%20binding"/>
> > >>
> > >>
> > >>Allen Day wrote:
> > >>
> > >>>This is a good region. I just realized the type ids coming back
in
> the
> > >>>/type response don't match the type ids given in the /feature
> response.
> > >>>I'm in the process of switching over from using table PKs for
types
> to
> > >>>using type names, so rather than using a filter "type=9770" you
can
> > >>>instead us a filter "type=sofa.ontology:CDS" (the "sofa.ontology"
> string
> > >>>is the term's namespace. SO has already changed it to "SOFA" in
> their
> > >>>most recent SOFA file in CVS, but I haven't updated yet).
> > >>>
> > >>>So, sorry about that. I just change thed /type response back to
give
> > >>>primary keys until I get names working. All the types should now
> work
> > >>>with the type= filter. Here's a quick document you can grab, a
> subregion
> > >>>of what you asked for in your request that has something
interesting
> > >>>inside:
> > >>>
> > >>>http://das.biopackages.net/das/genome/chado/chado-Hsa-
> 17/feature?overlaps=chr21/21292691:21324284;type=9688
> > >>>
> > >>>-Allen
> > >>>
> > >>>
> > >>>
> > >>>On Thu, 21 Apr 2005, Helt,Gregg wrote:
> > >>>
> > >>>
> > >>>
> > >>>>Cool, great to see a public server up and running!
> > >>>>
> > >>>>However, I'm having trouble with feature queries. For example:
> > >>>>http://das.biopackages.net/das/genome/chado/chado-Hsa-
> 17/feature?overlap
> > >>>>s=chr21/21500000:22500000;type=0000110
> > >>>>
> > >>>>Which is an example of the most typical query IGB will be making
--
> one
> > >>>>overlap filter combined with one type filter. But this query
> returns
> > >>>>with a single <FEATURELIST> line, no <FEATURES>. I've tried it
for
> > >>>>every type returned by the type query and get the same results.
> > >>>>
> > >>>>Is there something wrong with my query syntax?
> > >>>>
> > >>>>If this is not a good region to test, can you send me an example
> feature
> > >>>>query URL where you know what the response should be?
> > >>>>
> > >>>> thanks!
> > >>>> gregg
> > >>>>
> > >>>>
> > >>>>
> > >>>>>-----Original Message-----
> > >>>>>From: das2-bounces at portal.open-bio.org
> > >>>>
> > >>>>[mailto:das2-bounces at portal.open-
> > >>>>
> > >>>>
> > >>>>>bio.org] On Behalf Of Allen Day
> > >>>>>Sent: Wednesday, April 20, 2005 10:42 PM
> > >>>>>To: das2 at portal.open-bio.org
> > >>>>>Subject: [DAS2] Human prototype server up
> > >>>>>
> > >>>>>Contains UCSC Hg17 knownGene, as well as all Affymetrix
"target"
> > >>>>
> > >>>>sequences
> > >>>>
> > >>>>
> > >>>>>aligned to the genome w/ > 90% identity.
> > >>>>>
> > >>>>>http://das.biopackages.net/das/genome/chado/chado-Hsa-17/region
> > >>>>>
> > >>>>>Here's a URL to get you going. This is running on 2 hosts: a
dual
> > >>>>>opteron/8GB webserver frontend and a dual opteron/16GB database
> > >>>>
> > >>>>backend.
> > >>>>
> > >>>>
> > >>>>>There are no other applications on these hosts right now, so
hammer
> > >>>>
> > >>>>away.
> > >>>>
> > >>>>
> > >>>>>You might want to use HTTP 1.1 keepalive though, the responses
can
> be
> > >>>>>*really* slow for /feature requests on large regions. The
reason
> for
> > >>>>
> > >>>>this
> > >>>>
> > >>>>
> > >>>>>is the feature/part subpart requests are done feature by
feature
> due
> > >>>>
> > >>>>to
> > >>>>
> > >>>>
> > >>>>>the 1-many relationship... I'm still trying to figure out how
to
> > >>>>
> > >>>>improve
> > >>>>
> > >>>>
> > >>>>>this.
> > >>>>>
> > >>>>>Components that aren't working in even a basic way:
> > >>>>>
> > >>>>> * alignments
> > >>>>> * assemblies
> > >>>>> * individual /feature/<id> requests
> > >>>>> * versioned source response w/ supported formats
> > >>>>> * most feature filters (overlaps, type, and exacttype work)
> > >>>>>
> > >>>>>-Allen
> > >>>>>_______________________________________________
> > >>>>>DAS2 mailing list
> > >>>>>DAS2 at portal.open-bio.org
> > >>>>>http://portal.open-bio.org/mailman/listinfo/das2
> > >>>>
> > >>>_______________________________________________
> > >>>DAS2 mailing list
> > >>>DAS2 at portal.open-bio.org
> > >>>http://portal.open-bio.org/mailman/listinfo/das2
> > >>
> >
>
> On Thu, 21 Apr 2005, Ed Erwin wrote:
>
> > My reasoning is that it ISN'T a URL yet, so you don't escape it
until
> > you turn it into a URL. But Andrew can figure out which is right
and
> > document it.
> >
> > Allen Day wrote:
> > > exactly. my reasoning was that it was a relative url and needed
to be
> > > escaped.
> > >
> > > On Thu, 21 Apr 2005, Ed Erwin wrote:
> > >
> > >
> > >>This is a question for Andrew Dalke.
> > >>
> > >>
> > >>Allen Day wrote:
> > >>
> > >>>I don't know, is it? I have found the description of properties
and
> the
> > >>>/property response document examples to be extremely unclear. I
> opted to
> > >>>URI escape here, but maybe it's not the right thing to do.
> > >>>
> > >>>-Allen
> > >>>
> > >>>
> > >>>On Thu, 21 Apr 2005, Ed Erwin wrote:
> > >>>
> > >>>
> > >>>
> > >>>>That response comes back with some "%20" things in it. That
seems
> like
> > >>>>a bug to me, but is it?
> > >>>>
> > >>>> <PROP ptype="property/protein%20binding"/>
> > >>>>
> > >>>>
> > >>>>Allen Day wrote:
> > >>>>
> > >>>>
> > >>>>>This is a good region. I just realized the type ids coming
back in
> the
> > >>>>>/type response don't match the type ids given in the /feature
> response.
> > >>>>>I'm in the process of switching over from using table PKs for
types
> to
> > >>>>>using type names, so rather than using a filter "type=9770" you
can
> > >>>>>instead us a filter "type=sofa.ontology:CDS" (the
"sofa.ontology"
> string
> > >>>>>is the term's namespace. SO has already changed it to "SOFA"
in
> their
> > >>>>>most recent SOFA file in CVS, but I haven't updated yet).
> > >>>>>
> > >>>>>So, sorry about that. I just change thed /type response back
to
> give
> > >>>>>primary keys until I get names working. All the types should
now
> work
> > >>>>>with the type= filter. Here's a quick document you can grab, a
> subregion
> > >>>>>of what you asked for in your request that has something
> interesting
> > >>>>>inside:
> > >>>>>
> > >>>>>http://das.biopackages.net/das/genome/chado/chado-Hsa-
> 17/feature?overlaps=chr21/21292691:21324284;type=9688
> > >>>>>
> > >>>>>-Allen
> > >>>>>
> > >>>>>
> > >>>>>
> > >>>>>On Thu, 21 Apr 2005, Helt,Gregg wrote:
> > >>>>>
> > >>>>>
> > >>>>>
> > >>>>>
> > >>>>>>Cool, great to see a public server up and running!
> > >>>>>>
> > >>>>>>However, I'm having trouble with feature queries. For
example:
> > >>>>>>http://das.biopackages.net/das/genome/chado/chado-Hsa-
> 17/feature?overlap
> > >>>>>>s=chr21/21500000:22500000;type=0000110
> > >>>>>>
> > >>>>>>Which is an example of the most typical query IGB will be
making -
> - one
> > >>>>>>overlap filter combined with one type filter. But this query
> returns
> > >>>>>>with a single <FEATURELIST> line, no <FEATURES>. I've tried
it
> for
> > >>>>>>every type returned by the type query and get the same
results.
> > >>>>>>
> > >>>>>>Is there something wrong with my query syntax?
> > >>>>>>
> > >>>>>>If this is not a good region to test, can you send me an
example
> feature
> > >>>>>>query URL where you know what the response should be?
> > >>>>>>
> > >>>>>> thanks!
> > >>>>>> gregg
> > >>>>>>
> > >>>>>>
> > >>>>>>
> > >>>>>>
> > >>>>>>>-----Original Message-----
> > >>>>>>>From: das2-bounces at portal.open-bio.org
> > >>>>>>
> > >>>>>>[mailto:das2-bounces at portal.open-
> > >>>>>>
> > >>>>>>
> > >>>>>>
> > >>>>>>>bio.org] On Behalf Of Allen Day
> > >>>>>>>Sent: Wednesday, April 20, 2005 10:42 PM
> > >>>>>>>To: das2 at portal.open-bio.org
> > >>>>>>>Subject: [DAS2] Human prototype server up
> > >>>>>>>
> > >>>>>>>Contains UCSC Hg17 knownGene, as well as all Affymetrix
"target"
> > >>>>>>
> > >>>>>>sequences
> > >>>>>>
> > >>>>>>
> > >>>>>>
> > >>>>>>>aligned to the genome w/ > 90% identity.
> > >>>>>>>
> >
>>>>>>>http://das.biopackages.net/das/genome/chado/chado-Hsa-17/region
> > >>>>>>>
> > >>>>>>>Here's a URL to get you going. This is running on 2 hosts: a
> dual
> > >>>>>>>opteron/8GB webserver frontend and a dual opteron/16GB
database
> > >>>>>>
> > >>>>>>backend.
> > >>>>>>
> > >>>>>>
> > >>>>>>
> > >>>>>>>There are no other applications on these hosts right now, so
> hammer
> > >>>>>>
> > >>>>>>away.
> > >>>>>>
> > >>>>>>
> > >>>>>>
> > >>>>>>>You might want to use HTTP 1.1 keepalive though, the
responses
> can be
> > >>>>>>>*really* slow for /feature requests on large regions. The
reason
> for
> > >>>>>>
> > >>>>>>this
> > >>>>>>
> > >>>>>>
> > >>>>>>
> > >>>>>>>is the feature/part subpart requests are done feature by
feature
> due
> > >>>>>>
> > >>>>>>to
> > >>>>>>
> > >>>>>>
> > >>>>>>
> > >>>>>>>the 1-many relationship... I'm still trying to figure out how
to
> > >>>>>>
> > >>>>>>improve
> > >>>>>>
> > >>>>>>
> > >>>>>>
> > >>>>>>>this.
> > >>>>>>>
> > >>>>>>>Components that aren't working in even a basic way:
> > >>>>>>>
> > >>>>>>>* alignments
> > >>>>>>>* assemblies
> > >>>>>>>* individual /feature/<id> requests
> > >>>>>>>* versioned source response w/ supported formats
> > >>>>>>>* most feature filters (overlaps, type, and exacttype work)
> > >>>>>>>
> > >>>>>>>-Allen
> > >>>>>>>_______________________________________________
> > >>>>>>>DAS2 mailing list
> > >>>>>>>DAS2 at portal.open-bio.org
> > >>>>>>>http://portal.open-bio.org/mailman/listinfo/das2
> > >>>>>>
> > >>>>>_______________________________________________
> > >>>>>DAS2 mailing list
> > >>>>>DAS2 at portal.open-bio.org
> > >>>>>http://portal.open-bio.org/mailman/listinfo/das2
> > >>>>
> >
More information about the DAS2
mailing list