Perl 6 - the future is here, just unevenly distributed

IRC log for #opentreeoflife, 2014-07-29

| Channels | #opentreeoflife index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:07 josephwb joined #opentreeoflife
00:10 josephwb jimallman: you there?
00:10 jimallman josephwb: howdy
00:11 josephwb does the "Update MRCA" compare against the synthetic tree?
00:11 josephwb I thought it was taxonomy
00:11 jimallman i moved all that chatter about MRCA search to a proper GitHub issue:
00:11 jimallman https://github.com/OpenTreeOfLife/treemachine/issues/111
00:11 josephwb I saw
00:11 jimallman yes, it should be comparing against the latest reference (OT) taxonomy
00:11 jimallman sorry if I implied otherwise
00:11 josephwb oh, ok
00:12 josephwb i am looking at the emails for "Confusing results from getDraftTreeMRCAForNodes service"
00:12 jimallman … but the link i create opens the MRCA in the latest synthetic tree, which could be misleading i suppose.
00:12 kcranstn yes, I think that is misleading
00:12 kcranstn why would we not compare to the synthetic tree instead of the taxonomy
00:12 josephwb For curators, we want to compare against the taxonomy.
00:13 kcranstn for a naive user, that seems to imply that we consider the taxonomy to be the goal, not the synthetic tree
00:13 kcranstn gold standard = OTT, not tree
00:13 jimallman ah, perhaps i’m assuming wrong then. i’d need to check the code more closely
00:14 josephwb kcranstn: curators are working within some taxonomy (hopefully similar to OTT). Finding mislabelled taxa is easy against taxonomy.
00:14 kcranstn let me clarify, by “gold standard = OTT, not tree” I mean that we do NOT want to imply this
00:14 josephwb If synthetic tree breaks taxonomy, returned MRCA could be very different than what taxonomy would imply
00:15 jimallman kcranstn: you’re right, of course. it’s right in the name of the service, ‘getDraftTreeMRCAForNodes’
00:15 josephwb ok, i think we are arguing the same thing ;-)
00:15 kcranstn so the issue may be short term, while the synthetic tree quality is not very good
00:16 josephwb I see this as two separate things
00:16 josephwb 1) compare against taxonomy = troubleshooting problems with mapping, rooting, etc.
00:17 josephwb 2) compare against synthetic tree = some indication of conflict/agreement/other questions
00:17 josephwb I am fixated on curation at the moment
00:17 kcranstn ok, that makes sense to me
00:17 kcranstn just trying to make sure we aren’t confusing for new users
00:17 josephwb i think we could use both
00:17 kcranstn agreed
00:18 jimallman i thought perhaps there was a second method like getTaxoMRCA…  versus getDraftTreeMRCA… checking taxomachine now.
00:19 josephwb should be
00:19 josephwb somewhere
00:19 jimallman i see code that looks related, but no easy service (yet)
00:19 josephwb taxomachine should do what I want, treemachine against synthetic tree
00:19 josephwb i will loook at services
00:20 josephwb might need to get Cody involved, as things are not exactly clear with some things
00:20 josephwb (read: my limitations)
00:22 josephwb jimallman: where do you get the MRCA from?
00:22 josephwb in the curator
00:22 jimallman see the cURL calls I’ve been posting. those are literally how i test for MRCA in the curation app
00:23 josephwb *sorry, catching up *
00:23 jimallman found the similar method i was thinking of, it’s not a MRCA test but retrieves the taxonomic subtree from tips: extractTaxonomySubtreeForTipNodes
00:23 josephwb okay, so calling treemachine
00:24 jimallman yes. the cURL calls are in the GH issue: https://github.com/OpenTreeOfLife/treemachine/issues/111
00:24 jimallman typical calls, that is
00:24 josephwb this makes sense why some trees were not getting the expecetd MRCA
00:24 jimallman because we’re walking the synthetic tree, instead of the OT taxonomy?
00:24 josephwb e.g. the genus-level tree I sent earlier. If the genus is sunk, MRCA will be deeper.
00:24 josephwb YES
00:25 jimallman sunk = suppressed or ignored in the synthetic tree?
00:25 jimallman or doesn’t make it in due to lack of granularity?
00:25 josephwb no, I mean a genus is not monophyletic in the synthetic tree
00:25 jimallman ah, right
00:25 jimallman so it doesn’t “exist” there in a clear-cut way
00:25 josephwb exactly
00:26 josephwb so, i don't think anything is broke here.
00:26 josephwb just comparing against something other than I assumed
00:26 jimallman so it sounds like ‘extractTaxonomySubtreeForTipNodes’ might give a better result (albeit by working way too hard and getting the whole subtree)
00:27 josephwb yeah, would work much too hard
00:27 josephwb this is a taxomachine solution
00:27 jimallman hm, i just noticed that extractTaxonomySubtreeForTipNodes actually starts by setting the subtree root using our old pal getDraftTreeMRCAForNodes
00:27 josephwb i'll let Cody know
00:27 jimallman https://github.com/OpenTreeOfLife/treemachine/blob/248d6abd4367b17030afc8105b0258282c78eb36/src/main/java/opentree/GraphExplorer.java#L318
00:28 josephwb ah!
00:28 josephwb yes, let's stay away from treemachine for this
00:28 josephwb i image Cody could whip up the service very quickly
00:28 josephwb i am just not as familiar with his code base
00:29 jimallman please add a note to case 3 in issue #111 (link above), or i can do it if you prefer
00:29 jimallman we should add a comment i mean, and refer to “item 3 above”
00:31 josephwb *adding something*
00:48 kcranstn joined #opentreeoflife
01:03 towodo joined #opentreeoflife
01:52 josephwb joined #opentreeoflife
02:06 josephwb joined #opentreeoflife
04:07 josephwb joined #opentreeoflife
09:03 scrollback joined #opentreeoflife
11:08 josephwb joined #opentreeoflife
11:20 kcranstn joined #opentreeoflife
11:33 towodo joined #opentreeoflife
12:01 kcranstn joined #opentreeoflife
12:25 josephwb joined #opentreeoflife
13:50 kcranstn joined #opentreeoflife
14:03 kcranstn joined #opentreeoflife
14:08 towodo joined #opentreeoflife
14:17 towodo joined #opentreeoflife
16:00 towodo joined #opentreeoflife
16:03 jimallman joined #opentreeoflife
17:02 kcranstn joined #opentreeoflife
18:51 kcranstn joined #opentreeoflife
18:53 josephwb phy
18:53 josephwb oops, wrong window
18:54 kcranstn :)
21:31 kcranstn joined #opentreeoflife

| Channels | #opentreeoflife index | Today | | Search | Google Search | Plain-Text | summary