WPWG-20080930-transcript

From GO Wiki
Revision as of 12:32, 30 September 2008 by Sjcarbon (talk | contribs) (New page: <font color="#8D01A6"><font size="2">(08:32:03 AM) </font><font size="3">pascale:</font></font><font size="3"> je suppose que c'est une qualite</font><br> <font size="2">(08:32:26 AM) </fo...)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search
The printable version is no longer supported and may have rendering errors. Please update your browser bookmarks and please use the default browser print function instead.

(08:32:03 AM) pascale: je suppose que c'est une qualite
(08:32:26 AM) zenbitz [hitz@c-98-210-219-14.hsd1.ca.comcast.net] entered the room.
(08:32:32 AM) zenbitz is now known as ben
(08:32:35 AM) rama [rama@c-24-7-118-198.hsd1.ca.comcast.net] entered the room.
(08:32:39 AM) gwg: tres belle, et cetera... ;)
(08:33:31 AM) rama: hello
(08:34:06 AM) sjcarbon: morinin;
(08:34:48 AM) sjcarbon: lemme just pull up the wiki page...
(08:34:49 AM) cjm [cjm@adsl-76-202-119-204.dsl.pltn13.sbcglobal.net] entered the room.
(08:35:34 AM) cjm left the room (quit: Quit).
(08:36:01 AM) ben left the room (quit: Ping timeout).
(08:36:12 AM) sjcarbon: OK...
(08:36:24 AM) sjcarbon: <a href="http://ash.lbl.gov/cgi-bin/amigo/amigo">http://ash.lbl.gov/cgi-bin/amigo/amigo</a>
(08:36:31 AM) zenbitz [hitz@c-98-210-219-14.hsd1.ca.comcast.net] entered the room.
(08:36:42 AM) zenbitz is now known as ben
(08:36:45 AM) sjcarbon: Has everybody taken a look? <a href="http://ash.lbl.gov/cgi-bin/amigo/amigo">http://ash.lbl.gov/cgi-bin/amigo/amigo</a>
(08:36:52 AM) ben: hmm network trouble
(08:36:59 AM) ben: i mean with chat, not page
(08:37:22 AM) sjcarbon: I've added a few fixes that haven't appeared on ash yet.
(08:38:39 AM) sjcarbon: Also, I still need to run it through the w3 test stuff and fix the table styles a bit (done dirty).
(08:40:56 AM) sjcarbon: (Chris skyped me about some IRC problems.)
(08:41:08 AM) cjm [cjm@adsl-76-202-119-204.dsl.pltn13.sbcglobal.net] entered the room.
(08:41:32 AM) cjm left the room (quit: Quit).
(08:42:05 AM) cjm [cjm@adsl-76-202-119-204.dsl.pltn13.sbcglobal.net] entered the room.
(08:42:15 AM) cjm left the room (quit: Quit).
(08:42:28 AM) cjm [cjm@adsl-76-202-119-204.dsl.pltn13.sbcglobal.net] entered the room.
(08:44:23 AM) rama: seth
(08:44:40 AM) sjcarbon: hello!
(08:44:54 AM) rama: in the mean time please refresh our minds about what specific item is going to need solid testing
(08:45:09 AM) cjm2 [cjm@ftp.fruitfly.org] entered the room.
(08:45:12 AM) rama: there are lot of beta software on this 'ash' location
(08:45:33 AM) sjcarbon: Sorry--mostly beta to be ignored (for now).
(08:45:55 AM) sjcarbon: The Homolset Summary and the associated detail pages and graphs.
(08:46:08 AM) sjcarbon: <a href="http://ash.lbl.gov/cgi-bin/amigo/amigo?mode=homolset_summary">http://ash.lbl.gov/cgi-bin/amigo/amigo?mode=homolset_summary</a>
(08:46:11 AM) rama: okay thanks
(08:46:18 AM) sjcarbon: And links out from that page.
(08:46:33 AM) rama: are we really set on this big long table?
(08:46:46 AM) sjcarbon: Another way to break it up?
(08:47:00 AM) ben left the room (quit: Ping timeout).
(08:47:20 AM) rama: I know you have put lot of effort into this...but this is not very userfriendly
(08:47:47 AM) rama: thinking of workflow- what do you think an user will do after coming to this page?
(08:47:48 AM) pascale: why not? what do you think it should look like?
(08:48:23 AM) sjcarbon: Also remember that you can link directly from the gp-details pages to the ref genome details pages.
(08:48:32 AM) cjm2: There will be many ways that a user can get to a homolset summary - for example from a gene page
(08:48:42 AM) pascale: well in the long run it doesn't make so much sense to list *all* annotated genes, but given some selection cirteria, it's nice to be able to get some lists
(08:49:09 AM) gwg: are you going to zebra stripe the tables? It makes it a bit easier on the eye
(08:49:35 AM) sjcarbon: Still trying to figure out how without making the color thing too strange...
(08:49:36 AM) rama: for example ALG6
(08:49:55 AM) rama: there are 3 Zfin entries. It is confusing in my mind..
(08:50:24 AM) sjcarbon: yes, that is something that I'll have fixed later today...
(08:50:39 AM) rama: I understand the difference between the 3...but once you are at the bottom of the table you won't see the legend on the top and one will wonder
(08:51:04 AM) sjcarbon: i could add hover text...
(08:51:11 AM) sjcarbon: also, we were thinking about
(08:51:19 AM) sjcarbon: going back to an older version where
(08:51:35 AM) sjcarbon: there was just one symbol followed by information.
(08:53:06 AM) rama: I am not sure what is the best way to display this..we should send it out to curators and get some feedback. I could be in the minority
(08:53:21 AM) cjm2: rama - Val and I agree with you
(08:53:31 AM) zenbitz [hitz@c-98-210-219-14.hsd1.ca.comcast.net] entered the room.
(08:53:32 AM) rama: okay. thanks.
(08:53:34 AM) cjm2: see my email yesterday:
(08:53:43 AM) sjcarbon: It will be a fix for later today.
(08:53:44 AM) zenbitz is now known as ben
(08:53:50 AM) cjm2: -----
(08:53:51 AM) cjm2: I think it is really confusing to list the gene names multiply to indicate different types of evidence.
(08:53:51 AM) cjm2: I tend to agree
(08:53:51 AM) cjm2: Seth previously had this such the gene was only listed once. He will temporarily restore the previous display so we can compare.
(08:53:51 AM) cjm2: The confusion is compounded in the case of WB because they are providing annotations to both genes and gene products, meaning you see things such as eat-6 and EAT-6. Not a lot we can do about this until the data is fixed.
(08:53:52 AM) ben: I keep getting booted...
(08:53:53 AM) cjm2: ----
(08:54:06 AM) cjm2: (email snippet)
(08:54:06 AM) pascale: I wonder also how that table is sorted?
(08:54:29 AM) sjcarbon: That has been fixed, but is not showing up on ash yet.
(08:54:41 AM) sjcarbon: Alphabetical on my local (fixed) machine.
(08:54:57 AM) cjm2: Seth: I notice the sort option is now gone.
(08:55:22 AM) cjm2: Previously there was an option to sort by either high-information content terms or terms that are deep in the graph
(08:55:34 AM) cjm2: sorry ignore me
(08:55:44 AM) cjm2: that's on the individual homolset pages, doh!
(08:56:26 AM) ben: I suppose I don't need to point out that the images are gynormous, and don't always load properly.
(08:56:42 AM) sjcarbon: The SVGs or the PNGs?
(08:57:41 AM) rama: we talked about this before i think- about showing one ontology at a time.
(08:58:32 AM) sjcarbon: So for each set, six images? Also, won't the ontologies be blending soon?
(08:59:11 AM) rama: also rather than saying svg |png can we say ontology view or something explicit
(08:59:23 AM) rama: many people arent' familiar with what svg and png means
(08:59:41 AM) sjcarbon: yes, maybe interactive and static instead?
(08:59:42 AM) ben: both
(08:59:46 AM) cjm2: Seth - we can still opt to display each ontology individually even if wehave links between them
(09:00:11 AM) sjcarbon: ben - they work for me, can you give an example of one that fails for you?
(09:01:06 AM) cjm2: rama: what do you think are good labels? how about "static" image vs "navigable/zoomable" image?
(09:01:42 AM) ben: set 240/PNG- phenotype is that it drew a pic, spun and thrashed (I had a huge number of apps open, btw), then just didn't quite finish drawing.
(09:01:58 AM) ben: I did start force quitting stuff I wasn't using
(09:02:46 AM) cjm2: Perhaps there could be some warning for large images?
(09:02:53 AM) rama: how about - ontology view (static | dynamic)
(09:03:39 AM) sjcarbon: ben - the PNG itself seems fine, but it did take bunches of memory to initially load.
(09:03:47 AM) pascale: or graphical view?
(09:03:48 AM) sjcarbon: A warning would be good.
(09:03:57 AM) rama: graphical view works for me
(09:05:52 AM) rama: graphical view (static | zoomable/dynamic)- does that sound good?
(09:06:39 AM) sjcarbon: static vs. interactive makes it more obvious to me, but we can fiddle with the words as we go...
(09:06:50 AM) rama: i like interactive. nice.
(09:06:54 AM) pascale: me too
(09:07:35 AM) ben: I will reload the image and see.
(09:07:50 AM) sjcarbon: Thanks.
(09:08:50 AM) ben: I have a 2x2Ghz intel mac, with 1.5Gb of RAM
(09:09:17 AM) gwg: monitor size?
(09:09:19 AM) ben: PNG might eventually load... but I think most users are going call "fail" after a little while.
(09:09:23 AM) sjcarbon: Sounds more or less like mine.
(09:09:49 AM) sjcarbon: Hmmm...I could try shrinking them a bit.
(09:09:55 AM) ben: it's a macbook pro, what like 13"? Not the huge one.
(09:10:04 AM) ben: Lower the resolution?
(09:10:29 AM) ben: this pic is 12735x7173
(09:10:29 AM) sjcarbon: Yeah, that should get the bit count down.
(09:11:08 AM) sjcarbon: Maybe try and see if I can get other things limited as well...
(09:11:23 AM) ben: it is a 15", 1440x900. I guess the big one is 17"
(09:11:36 AM) sjcarbon: I'll see what I can do and get back to the group.
(09:12:46 AM) cjm2: Some indication of size of the graph would be useful also to given an indication of how widely distributed annotations are for particular homolsets
(09:13:13 AM) sjcarbon: a node count?
(09:13:31 AM) cjm2: Would it be overloading the table too much to show the total number of annotated nodes (incl transitive) for each homolset?
(09:13:50 AM) sjcarbon: That was something people hated previously.
(09:13:52 AM) ben: ok , it stopped loading after about 5 min, still has the bottom part of the graph cut off.
(09:13:59 AM) sjcarbon: Safari?
(09:14:07 AM) ben: FF2.x+
(09:14:29 AM) cjm2: ok. I tend towards information-rich displays.. perhaps this could be a power user option
(09:14:42 AM) sjcarbon: Ben - I don't suppose I could talk you into FF3+...?
(09:14:55 AM) ben: 2.0.0.17
(09:15:13 AM) sjcarbon: Cool.
(09:15:24 AM) ben: I will switch eventually.
(09:15:26 AM) pascale: but some users may have V2 still
(09:15:27 AM) sjcarbon: I'll see if there are specific issues for that as well.
(09:15:30 AM) gwg: but you won't be able to talk all the users into FF3+
(09:15:45 AM) gwg: the functionality should not be browser dependent
(09:15:54 AM) sjcarbon: Not the PNG stuff at least.
(09:16:14 AM) gwg: there may even be users with (whisper it!) IE! :o
(09:16:26 AM) ben: I would vote for a cytoscape webstart
(09:16:42 AM) ben: sure it's a 10Mb download....
(09:17:23 AM) ben: It's not that hard to do, although they have bumped a couple versions since last I tried.
(09:20:23 AM) rama: we are looking at cytoscape webstart for the png stuff?
(09:20:24 AM) sjcarbon: Is there anything else that people want to talk about before the next iteration? I'll try and see how much ground I can cover today.
(09:20:32 AM) rama: couple of questions
(09:20:50 AM) rama: 1) are you going to demo the new ref.genome views at the GOC?
(09:21:03 AM) rama: 2) are curators going to test it before the GOC?
(09:21:18 AM) sjcarbon: If you think it's test-worth, then we can.
(09:21:27 AM) sjcarbon: It would be nice...
(09:22:07 AM) rama: I think it would be useful to get the attention of the larger GOC
(09:22:34 AM) rama: because it seems like people are working on their own thing and would love to see what is going on
(09:23:08 AM) rama: it is test-worthy if you can revert it to displaying the gene name once
(09:23:23 AM) sjcarbon: one of today's projects!
(09:23:46 AM) sjcarbon: I'll see how much I can get done today, and send the URL out tonight.
(09:23:48 AM) rama: I would definitely send it to ref-genome/annotation mailing lists or just the GO mailing list
(09:23:49 AM) cjm2: great! once it's done should we send it to a wider group or the whole GO list?
(09:24:11 AM) sjcarbon: I think the (ancient) initial plan was the wider group.
(09:24:17 AM) cjm2: ok
(09:24:27 AM) sjcarbon: Although now the testing machine might be somewhat underpowered...
(09:24:28 AM) rama: ref.genome and annotations mailing list shd do
(09:24:53 AM) rama: please give specific time line for testing (please don't leave it open ended)
(09:25:06 AM) pascale: right!
(09:25:09 AM) rama: and specific questions will help
(09:25:14 AM) sjcarbon: Will do.
(09:25:19 AM) pascale: and ref genome and annotation lists are more likely to look at it seriously
(09:25:25 AM) rama: thanks
(09:25:37 AM) pascale: although it woud be good to send to the wider group as well
(09:25:48 AM) sjcarbon: Anything else today?
(09:26:02 AM) pascale: not for me
(09:26:07 AM) rama: i think we should give folks on these two mailing list 1 week time, fix bugs and then send it to wider list
(09:26:17 AM) sjcarbon: sounds good.
(09:26:22 AM) rama: what happened to the dual taxon id
(09:26:33 AM) gwg: nothing as yet
(09:26:48 AM) rama: I saw a posting by Michelle over the weekend and was reminded of that
(09:26:57 AM) ben: we no longer ignore most of them!
(09:27:06 AM) ben: sw group is going to talk about it tomorrow
(09:27:16 AM) ben: can't get them in the Db at the moment.
(09:27:31 AM) rama: okay. thanks.
(09:27:43 AM) sjcarbon: Anything else?
(09:27:48 AM) rama: they are in the DB i think
(09:28:05 AM) rama: i don't know for sure
(09:28:33 AM) sjcarbon: Going once...
(09:28:43 AM) sjcarbon: Going twice...
(09:28:43 AM) rama: Action item?
(09:28:47 AM) sjcarbon: Yes?
(09:29:38 AM) rama: send the ref. genome stuff for testing to the 2 mailing lists
(09:29:46 AM) rama: thats all from me. thanks all. bye
(09:29:57 AM) pascale: bye
(09:29:59 AM) sjcarbon: After y'all take another look. Yes. Added.
(09:30:00 AM) pascale left the room.
(09:30:02 AM) sjcarbon: Bye!
(09:30:07 AM) rama: bye
(09:30:10 AM) rama left the room (quit: Quit: Computer goes to sleep!).