[DAS] Link Command

Dave Wall dawall@us.ibm.com
Tue, 1 Oct 2002 10:30:23 -0500


That makes some sense.  The servers are returning an empty record so it
sounds like that is what they are supposed to do.  That means the
information I am looking for is in the results of the 'feature' command(?).

Does an example scenario (using the wormbase server) become?
   http://www.wormbase.org/db/das/dsn.  Returns the list of data sources on
   this server.  The only server that is returned is 'elegans'
   http://www.wormbase.org/db/das/elegans/entry_points.  Returns a list of
   entry points.  For the purpose of this discussion I will select
   <SEGMENT id="X" size="17683571" start="1" stop="17683571" class="
   Sequence" orientation="+" subparts="yes">X
   </SEGMENT>
   http://www.wormbase.org/db/das/elegans/features?segment=X:1,100000.
   Returns a list of features against this segment.  Many feature records
   are returned (around 300).  An example record is:
   <FEATURE id="Y73B3A.1" label="Y73B3A.1">
              <TYPE id="exon" category="transcription" reference="no">
            curated exon</TYPE>
              <METHOD id="curated">curated</METHOD>
              <START>24479</START>
              <END>24819</END>
              <SCORE>-</SCORE>
              <ORIENTATION>-</ORIENTATION>
              <PHASE>0</PHASE>
              <GROUP id="Y73B3A.1">
                <LINK href="
              http://www.wormbase.org/db/get?name=Y73B3A.1;class=Sequence">
              Y73B3A.1</LINK>
            </GROUP>
          </FEATURE>


        I see the "link" which is part of the <group>.  Is that the
        annotation or does this entry not have one since there is no <link>
        under the feature tag?  Actually none of the entries have a <link>
        parameter under the feature tag.  If one exists it is in the
        <group> section.  Am I getting closer?

        Thanks again!



David Wall
553-5329
IBM Life Sciences



                                                                                                                          
                      Lincoln Stein                                                                                       
                      <lstein@cshl.org>        To:       Dave Wall/Rochester/IBM@IBMUS, das@biodas.org                    
                                               cc:                                                                        
                      10/01/2002 06:32         Subject:  Re: [DAS] Link Command                                           
                      AM                                                                                                  
                      Please respond to                                                                                   
                      lstein                                                                                              
                                                                                                                          
                                                                                                                          



You're not going crazy.  It's just that all the servers to date have chosen
to
implement the client-side form of link (where the link information is
returned gratis with every feature record) rather than implement
server-side
link.  The servers should be returning an empty record, are they?

Lincoln

On Friday 27 September 2002 11:46 am, Dave Wall wrote:
> Geoff and I are stumped.  What I am looking for is a working example of
the
> link command.  I have other commands working but I cannot get the link
> command to return an annotation.  I have tried a couple different severs
> (Geoff tried others) but no success yet.  An example is
>
>
http://genome.ucsc.edu/cgi-bin/das/hg12/link?field=feature;id=GC.chr18.6000
>0
>
> It may be that my syntax is wrong, or maybe my whole approach is wrong.
Am
> I going to the wrong server?  Is my use of the link command incorrect?
> Thanks for any help!
>
> David Wall
> 553-5329
> IBM Life Sciences
>
> ----- Forwarded by Dave Wall/Rochester/IBM on 09/27/2002 10:38 AM -----
>
>                       Geoff Purdy
>                       <gpurdy@cogentneuros        To:       Dave
> Wall/Rochester/IBM@IBMUS cience.com>                 cc:
>                                                   Subject:  RE: [DAS]
> Getting Started With DAS 09/27/2002 09:36 AM
>
>
>
>
>
> Well, I'm out of ideas.  There is evidence that "link" works on the
Sanger
> DAS server as the pop-up menus are populated with additional information.
> However, I've been unable to access that information using a script.
>
> I suggest posting your question on the biodas list.  The collective
> expertise on the list is orders of magnitude greater than my own.
>
> I don't know of any alternatives to "link" for accessing annotation
> details.
> The reason I quoted "link" is that command names appear in italics in the
> DAS documentation and italics isn't available in a plain text message.
>
> > -----Original Message-----
> > From: Dave Wall [mailto:dawall@us.ibm.com]
> > Sent: Thursday, September 26, 2002 5:29 PM
> > To: Geoff Purdy
> > Subject: RE: [DAS] Getting Started With DAS
> >
> >
> >
> > Thanks for looking into it.  If I read between the lines a little it
>
> almost
>
> > sounds as if there is another way to get the annotation.  You  tried
>
> three
>
> > servers and none of them delivered the data via the link command.  Is
>
> there
>
> > another way to do it?
> >
> > David Wall
> > 553-5329
> > IBM Life Sciences
>
> Dave,
> You raise a good point here.  I notice that if I use the link command on
> three different das servers, I see three different behaviors.
>
> Last I saw, the UCSC DAS server was Jim Kent's C implementation of the
DAS
> specification.  As you pointed out, the link command gives no output (in
> the
> handful of cases I've tried).
>
> I also tried a similar query on the Sanger Center DAS server (a Java
> implementation, I believe).  This server gives the following message if a
> feature doesn't have associated annotations:
>    <h1>No extra information about this object</h1>
>    <p>field: feature<br />id: 18/p11.32</p>
>
> Finally, I tried CSHL's wormbase DAS server (a perl implementation by
> Lincoln Stein).  This doesn't seem to implement the link command at all.
>
> I notice that the "link" command is an annotation server specific command
> (not part of reference server specification).  I'm unsure which of these
> servers purport to be reference, annotation, or both.  Another
possibility
> is that different implementations of DAS may omit implementing portions
of
> the specification.
>
> I'll give this a closer look, but I'm beginning to think this might be a
> good subject for on-list discussion.
>
> > -----Original Message-----
> > From: Dave Wall [mailto:dawall@us.ibm.com]
> > Sent: Thursday, September 26, 2002 4:15 PM
> > To: Geoff Purdy
> > Subject: RE: [DAS] Getting Started With DAS
> >
> >
> >
> > Thanks for offering to help!
> >
> > I have been having pretty good luck getting a list of data
> > sources, types,
> > segments, etc., but crash when I trying to get an annotation.
> >  Here is an
> > example for the purpose of verifying URL syntax ...
> >
> >
> > http://genome.cse.ucsc.edu:80/cgi-bin/das/hg12/features?segmen
> > t=18:1,100000
> >
> > In the returned data is
> >       - <FEATURE id="GC.chr18.60000" label="GC">
> >               <TYPE id="gcPercent" category="other" reference="no">GC
> >             Percent</TYPE>
> >               <METHOD />
> >               <START>60001</START>
> >               <END>80000</END>
> >               <SCORE>0</SCORE>
> >               <ORIENTATION>0</ORIENTATION>
> >               <PHASE>-</PHASE>
> >             - <GROUP id="GC.chr18.60000">
> >                 <LINK href="
> >
> > http://genome.ucsc.edu/cgi-bin/hgTracks?position=chr18:60000-8
> > 0000&db=hg12
> >               ">Link to UCSC Browser</LINK>
> >             </GROUP>
> >           </FEATURE>
> >
> > I then send the following which returns a blank page.
> >
> > http://genome.cse.ucsc.edu/cgi-bin/das/hg12/link?field=feature
> > ;id=GC.chr18.60000
> >
> > I am probably missing something obvious, but I sure can't
> > figure out what
> > it is.  Thanks again!
> >
> > David Wall
> > 553-5329
> > IBM Life Sciences
>
> _______________________________________________
> DAS mailing list
> DAS@biodas.org
> http://biodas.org/mailman/listinfo/das

--
Lincoln Stein
lstein@cshl.org