[DAS2] Notes from the weekly DAS/2 teleconference, 9 Oct 2006

[Steve Chervitz]

Notes from the weekly DAS/2 teleconference, 9 Oct 2006

$Id: das2-teleconf-2006-10-09.txt,v 1.1 2006/10/09 17:24:14 sac Exp $

Note taker: Steve Chervitz

Attendees: 
  Affy: Steve Chervitz, Ed Erwin, Gregg Helt
  UCLA: Allen Day, Brian O'connor

Action items are flagged with '[A]'.

These notes are checked into the biodas.org CVS repository at
das/das2/notes/2006. Instructions on how to access this
repository are at http://biodas.org

DISCLAIMER: 
The note taker aims for completeness and accuracy, but these goals are
not always achievable, given the desire to get the notes out with a
rapid turnaround. So don't consider these notes as complete minutes
from the meeting, but rather abbreviated, summarized versions of what
was discussed. There may be errors of commission and omission.
Participants are welcome to post comments and/or corrections to these
as they see fit. 


Agenda
-------
* Status reports


Topic: Status reports
---------------------
gh: Funding thru end of may. shifting times around a bit here at
affy. gh going up to a greater percentage during this period.
going down to half time for next month due to house-related work.

Focusing now on cleaning up impl of writeback on igb client. clean
impl based on ideas sketched out at code sprint in Aug.

Spec issue:
-----------
gh: was there a resolution to the feature group assembly conversation
on email thread. 
aday: died out. so the assumption is: no change.

[A] Ask andrew about feature group assembly resolution, if any.


ee: new release of IGB. bug fix then patch release. rapid turn
around. Exposed need for more throurough testing.
Specifying multiple urls for get more info links. sources for urls:
track lines in psl/bed files. Also supporting das files (1 and
probably 2)
noticed: feature tag can give feat label and ID. IGB ignores these
labels, because they seem to be attached to wrong thing. feat in das/1
is like 'exon' group is 'mrna'. it's the mrna we want the label on,
not exon where the labels are on.

gh: if people just label parent. names don't have to be unique. id is
unique uri, name is displayed name. parser isn't looking into that now.

[A] Ed will look into using feature name as label in IGB client


sc: Installed updated das2_server code on affy the das/2 server
(netaffxdas.affymetrix.com). Installed new, efficient version of exon
array data for hg18 (Mar 2006) assembly on this server. (igb's 'Bprobe1'
parser, generates new bp2 format files). Probe and probeset data
loaded fine, but exon/transcript cluster data failed with exception
about 'Probe_count is zero for <probesetID>'.

gh: problem: the bp2 data format isn't designed for representing
transcripts/exon just probe. problem in the part that generates the
bp2 files. can take a look at that.

[A] Gregg will look into steve's Bprobe1 parser error. Needs source gff.

ee: Can you verify that the gff data you are loading doesn't have
unmapped probes, probe sets? Some are not mapped after lifting from
previous genome assembly.

[A] Steve will remove unmapped objects in the source gff used for bp2


aday: working on UML for integrating the writeback and the read
features. Also retrieval of dynamic features as well. Sent out example
query. working on getting them all into a single model, determines
what do do based on input query.

will impl own block caching rather than apache caching.
If I see a writeback coming in , can see which types have been
modified, within each region. can fork off process to re-generate them
after doing the writeback. will be a lot faster.

Have a flowchart. partway through creating UML classes, functions,
return types. Using poseidon.

[A] Allen will distribute uml diagrams for das/2 modeling when ready

gh: will locking be a part of that?
aday: can make sure it's compatible. don't know how much of that to
impl now.
gh: useful to think about how to model that too.

[A] Allen will include locking in his UML modelling.

aday: flowchart is pretty generic. can be used by other servers.


bo: no das work because of work on manuscript.
started sourceforge project for das/2 assay "gyrax" (nee hyrax --
already taken at sf).
The motivation for this project is to take the das/2 objects in igb
and make them more generic. This project can host these objects. They
could then be used for other apps (igb, gyrax, others). Mark
Carlson in lab is working on the gyrax client.  Could be a nice
library for use by other apps, gui or not, that are built on top of a
das server.

gh: parts of the igb objects are tied into genometry model, a separate
package also. but both of these could be separated from igb.

ee: There was some email on genoviz forum where someone is writing an
app based on old NGSDK objects, on the help forum on
sourceforge. problems with >30,000 glyphs. advice: switch to efficient
glyph versions (special drawing alg if children are too small to see).

gh: Lots of caveats...There is code that hasn't been touched in a
while.

gh: question about hardware quote for UCLA

[A] Allen will send gregg hardware quote for UCLA (<$5k)

sc: status of hardware for affy das server upgrade?

gh: plan to order end of oct, should have in place in first two weeks
of nov.