2008-10-31 Project Wide Conf Call
Attending: Ilya, Kevin, Curtis, Melissa, Andrew, Chris, Jean-Marie, Brian, Jason, Aleksandra, Donald, Will, Josh
Agenda
Next Data Model release.
Other Data Model - Is replicates messing from SPW?
The query concerns ScreenPlateWell. The experimentalists perform screens using physical plates that were spotted by a robot according to a pre-determined pattern. A batch of each design or layout is made in one go and stored stored in the freezer for later use. The biologist then requests a selection of plates to cover the needs of their experiment, using the replicates to deal with defects and increase statistical accuracy of their measurements.
So, we have a Screen concept consisting of M different Plate layouts, each Plate layout being replicated N times, so there are M*N physical plates in total (see attached diagram). The issue is how to model the abstract plate layouts (each has a home-made catalogue number) and also the physical plate replicates.
Currently, as I understand it, there's no abstract concept of a plate layout. Also, to store the replicate number, one would have to let the Plate.Name attribute (or perhaps Plate.Status) contain a compound identifier (eg., Name in 1.1, 1.2, 1.3, ..., 2.1, 2.2, ...), but this obviously nasty as the structure isn't explicit.
Perhaps, there could be a PlateLayout entity, or similar, that describes the contents of the wells, referenced by Plate. Then Plate would be extended to include {Plate.Name, Plate.Replicate} as a pair of secondary keys.
Supporting OME Server
OMERO iterations
- Definition and release schedule for Beta3.2
- Metadata in Beta3.2
- Testing OMERO.web
ASCB (booth, travel, equipiment, what we want to show)
ASCB Poster - improve and update SBS version view here
Meeting page instructions
Minutes
Meetings
- this time on Fridays?
- Works great (Kevin+Ilya)
- Be sure to take agenda seriously. Everyone has access to plone.
-
Andrew looking to writing the instructions on meetings.
- Technology is still a bit unclear.
- E.g. Skype would be nice if it worked.
- Curtis & Ilya had it working, Hopefully Dundee next time.
Schema
- release coming up? But does it make sense to hold off.
-
Release as soon as possible? Was a decision made about bindata, tiffdata arity?
- Change made to model rather than to the code-generation. As Curtis wanted it. Only one or the other.
- If a third type is added, we can re-evaluate.
- Ok. Andrew sends out emails this weekend with list of changes.
- Any other schema questions?
Request from Nigel (Heidelberg) re: screens
- Wants an additional layer between plates and wells
- His model is strictly hierarchical with plates within screens.
-
Ilya: You certainly have plates in imaging which are replicates
- But doesn't the model accomadate that with the reagent plates
- They're disconnected from the imaging plates.
- Just not a hierarchical structure for just that reason.
- Chris: Problem may be due to nomenclature. "What's a plate?"
- Need to come up with a way for representing what Nigel wants.
- Perhaps small diagrams of various workflows in this model would help.
-
Jason: Similar to Andrew, people are contacting us privately,
- but all the communication should be pushed toward the user list.
- Good for the community to see everything that's going on,
- even if they don't "reply to all"
OME server questions
- Bandwidth on Baltimore's side?
-
Context? Several people have been writing to the list over the last few weeks.
- Some new installs
- Some people trying to use OMERO clients against OME-Perl
- Loci doesn't currently have an OME-Perl server to verify issues.
- No one is dedicated to OME-Perl other than Ilya. Ad currently out.
- Idea was to do best effort.
-
New release from HEAD. Backburner project.
- Database changes, analysis engine, semantic types, internationalization
- No timeline at the moment.
- Put WND-CHARM on next agenda? Sure.
OMERO
-
3.2 release. Things to fix everywhere.
- Server: Thumbnails, Sessions, Postgres versions
- Importer: LSID fixes, file formats, launcher with comment box (codec jar)
-
Insight: problems with textual annotations (with history of moves)
- Issue with multiple ports for OMEROs on same machine. (Beta4 feature)
- OmeroJava updated... (inaudible)
-
When?
- Chris needs time to work on thumbnailing
- Donald needs to look at some stuff
- J-M has weird Windows bugs
- Freeze metadata cycle? And focus starting Monday on release?
- Merge on Monday? Everything breaks?
- Jason: Hesitant to break that momentum.
- Are the bugs critical? Yes. * Leica is most dominant format and it's broken. Etc.
- Fixing archive original file for Beta3.2? No. * Turn it back on with warning. * Sounds really, really! slow. (Mainly Leica)
- Need to break all clients to make progress with Metadata
- Decision: Monday stopping metadata and fixing bugs.
-
3.2 release. Things to fix everywhere.
Bioformats
- Try to go to ASCB with 6 of the formats as complete as possible.
-
What's the best way to make this obvious? Have people bring data to ASCB?
- Visible in ImageJ, OMERO, ...
- "6 down, 59 to go"
- Some way to get the point across.
-
Kevin: poster/banner just for bio-formats
- Agreed. We need these ideas.
- Melissa + Poster == Good.
- Corner booth so two open sides.
Webclient
- Testing version and logins available. Will set up a proxy for logins
ASCB
- Andrew is looking into AV materials for hire
-
...what are we going to demo & what machines/servers do we need to have (running)?
- How many plugs? Omero server, Plasma monitor machine, 3 laptops at tables (8ft)?
- In a good location this year.
- Might make sense to have on loop bioformats opening up files and showing metadata
- Lots of people didn't have questions for developers, but were just looking.
- Kevin will bring vinyl banners.
-
Something on outer corner of booth
- Table with drape? Tall black (display) pillars?
- Coffee? Costing is unbelievable. 1000USD+
Josh loses connectivity...