Perl 6 - the future is here, just unevenly distributed

IRC log for #opentreeoflife, 2014-06-17

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

All times shown according to UTC.

Time Nick Message
00:00 josephwb joined #opentreeoflife
01:01 josephwb joined #opentreeoflife
01:39 jimallman joined #opentreeoflife
02:00 mtholder joined #opentreeoflife
02:05 mtholder joined #opentreeoflife
03:08 jimallman mtholder: any idea why the Bibliographic References list would be empty on devtree?
03:08 jimallman http://devtree.opentreeoflife.org/about/references
03:09 jimallman I just pushed up some changes, but neglected to test this list beforehand. my new version works in a local test setup.
03:09 jimallman (finds a big ol’ list of contributing trees)
03:09 jimallman i’m guessing oti or something needs to be re-indexed…?
03:11 josephwb joined #opentreeoflife
03:16 jimallman josephwb: do you know if oti or another index is borked in devtree?
03:16 jimallman i’m trying to figure out whether i’ve broken the Bibliographic References list:
03:16 jimallman http://devtree.opentreeoflife.org/about/references
03:19 josephwb sorry jimallman, no idea
03:20 josephwb "borked"?
03:20 jimallman ok, thanks. i’m going to chill until morning and see if the list re-appears..
03:20 jimallman http://www.urbandictionary.com/define.php?term=borked
03:21 josephwb hah, just came upon that myself
03:21 josephwb i am so unhip...
03:22 jimallman i picked it up in nerdly circles, long ago. see also cruft.
03:22 josephwb yes, that was also new for me today
03:22 josephwb truth be told, i think cody was just going along with it
03:22 jimallman :D
03:22 josephwb without knowing *exactly* what it meant
03:22 josephwb me too
03:23 josephwb context fills in a lot
03:23 jimallman it’s a great word! (perfectly cromulent, as they say)
03:24 josephwb jimallman, you are a bona fide frood
03:24 jimallman uh-oh, that’s vaguely familiar…
03:24 josephwb what?!?
03:24 josephwb you know every reference ever!
03:24 jimallman right, hoopy frood! thanks
03:25 josephwb how do you not instantly recall the ultimate reference?
03:25 josephwb *phew*
03:25 jimallman my memory is … weird.
03:25 jimallman more of a coral reef than a codex. some things stick, others don’t.
03:26 josephwb *urge to kill fading, fading…*
03:26 josephwb *rising*
03:26 josephwb *fading…*
03:26 jimallman wow, another one i did not know. awesome.
03:27 jimallman *gone*
03:28 josephwb sorry i can't help you out on the oti thing
03:29 jimallman no sweat. we’ll see what the morning brings. i think i’m done for the night!
03:29 josephwb l8r
03:30 josephwb this chat is cruft anyway ;-)
03:32 josephwb joined #opentreeoflife
08:21 jimallman joined #opentreeoflife
08:21 17SAAC7EF joined #opentreeoflife
08:44 jimallman joined #opentreeoflife
08:44 17SAAC7EF joined #opentreeoflife
10:35 towodo joined #opentreeoflife
10:52 josephwb joined #opentreeoflife
11:30 josephwb joined #opentreeoflife
12:20 josephwb joined #opentreeoflife
12:23 mtholder joined #opentreeoflife
12:55 mtholder joined #opentreeoflife
13:06 josephwb r u there jimallman?
13:07 josephwb maybe this isn't just for jimallman then
13:07 josephwb working on the R package, using curl calls
13:08 josephwb at some points, "api.opentreeoflife" works
13:08 josephwb other times, "devapi.opentreeoflife"
13:08 josephwb and still others "dev.opentreeoflife"
13:09 josephwb have we stabilized to something?
13:09 josephwb probably "api.opentreeoflife", right?
13:11 jimallman josephwb: hi, here now
13:11 josephwb yeah, so using curl from within R
13:12 josephwb but i have to keep tweaking it depending on which server we are using
13:13 josephwb "http://api.opentreeoflife.org/treemachine/v1/getSyntheticTree" is currently working
13:13 josephwb "api.opentree…" will be the final one, right?
13:13 jimallman hm, i believe api (ot12) and devapi (ot10) are both live
13:13 josephwb what is the difference?
13:13 jimallman and should have the same methods available (probably newer on devapi)
13:14 jimallman ot12 is frozen for the hackathon, ot10 has the latest changes
13:14 josephwb devapi not working for me
13:14 jimallman weird results? or not responding at all?
13:14 josephwb Error: Internal Server Error
13:15 josephwb that is an R error
13:15 josephwb let me try from the commandline
13:15 jimallman ah, OK. fyi, devapi pings for me, and this call works (but gives an empty result):
13:15 jimallman curl -X POST http://devapi.opentreeoflife.org/treemachine/ext/GoLS/graphdb/getSynthesisSourceList
13:15 jimallman (that’s why my Bibliograpic References page is failing)
13:16 josephwb josephwb$ curl -X POST http://devapi.opentreeoflife.org/treemachine/v1/getSyntheticTree -H "content-typ:application/json" -d '{"format":"newick", "maxDepth":"3", "subtreeNodeID":"3534540"}'
13:16 josephwb {
13:16 josephwb "message" : "Node 3534540 not found",
13:16 josephwb "exception" : "NotFoundException",
13:16 josephwb "fullname" : "org.neo4j.graphdb.NotFoundException",
13:16 josephwb "stacktrace" : [ "org.neo4j.kernel.impl.core.NodeManager.getNodeById(NodeManager.java:339)", "org.neo4j.kernel.InternalAbstractGraphDatabase.getNodeById(InternalAbstractGraphDatabase.java:959)", "opentree.GraphDatabaseAgent.getNodeById(GraphDatabaseAgent.java:89)", "opentree.GraphExplorer.reconstructSyntheticTree(GraphExplorer.java:2460)", "opentree.plugins.GoLS.getSyntheticTree(GoLS.java:372)", "java.lang.reflect.Method.invoke(
13:16 josephwb sorry so messy
13:17 josephwb doing the same call without "dev" works
13:17 josephwb why isn't the node being found?
13:19 jimallman i suspect the treemachine data on this server needs to be reloaded..
13:20 josephwb sounds right
13:20 jimallman mtholder: towodo: any ideas why the treemachine data on devapi (ot10) would be empty or out of date? do we need to reload from somewhere?
13:21 * jimallman is grabbing coffee, brb
13:35 blackrim joined #opentreeoflife
13:39 blackrim jimallman: hm, seems like this page is still not updated with the changes that I pushed. I thought you were doing something that would update it. did i do something wrong?
13:40 jimallman sorry, which page?
13:40 blackrim i missed the paste there developer-resources
13:40 jimallman Dev Resources in the main webapp?
13:40 blackrim yeah
13:40 jimallman let’s see...
13:41 josephwb blackrim: think I found the problem in the synth tree (re: ':')
13:41 jimallman i’m guessing that devtree is good, but production is stale
13:41 blackrim josephwb: ok, good
13:41 jimallman http://devtree.opentreeoflife.org/about/developer-resources
13:42 jimallman http://dev.opentreeoflife.org/about/developer-resources
13:42 blackrim jimallman: aha, yeah, that one is good.
13:42 blackrim are we not updating the production one?
13:44 jimallman it’s because production (confusing called “dev”) is running on the ‘production’ branch of opentree, while devtree uses the ‘master’ branch. we need merge some safe and favored changes from master to production, then re-deploy.
13:45 blackrim ok, that is fine. i just corrected some of the links on that one page but it is fine to wait (assuming those will update sometime in the near future)
13:59 towodo joined #opentreeoflife
14:09 kcranstn Sounds like we want to set up a deployment schedule for production
14:13 towodo yes, I was thinking along those lines. like, set a date, and say we're going to revise tree. and api. on that date? (with curator, unique api, etc.)
14:15 josephwb are we letting BOLD names through?
14:15 josephwb were we always?
14:16 josephwb getting a weird newick error that never arose before
14:16 josephwb names containing colons
14:16 josephwb e.g. Genus_sp._BOLD:XXXX
14:17 josephwb e.g. Genus_sp._BOLD:XXXX_otNNNNN
14:17 towodo OTT 2.8 has a very new NCBI in it.  That could easily have colons in names.
14:17 josephwb ok
14:17 josephwb another question: we want these sp. taxa?
14:18 towodo oh yeah.  i don't think so.  i realized recently I hadn't suppressed them, even though it had been decided they should be
14:18 towodo meant to fix that.
14:18 josephwb phew, okay, thought i missed something. thanks.
14:20 josephwb still need to accommodate possible semicolons, so this is good to know
14:20 towodo I think sp. affects about 80 OTUs used in studies
14:21 josephwb a "sp." in a tree is fine (we prune the "sp." and map to the higher taxon)
14:22 josephwb but it shouldn't be in the taxonomy (right?)
14:22 towodo correction: 'sp.' taxa are fine, if there's something after the 'sp.'
14:23 towodo it's the ones whose names end in 'sp.' that are trouble
14:23 towodo "\\bsp\\.$"    get flagged not_otu
14:24 towodo josephwb^
14:24 josephwb the BOLD "taxa" seem a lot like the whole strain problem
14:24 josephwb e.g. 500 BOLD sequences might all come from a robin
14:25 josephwb 500 separate robins, of course (or, probably)
14:25 towodo yes. and there are gajillions of them.
14:25 josephwb I vote: *chuck 'em*
14:25 towodo we could filter them out ad hoc, based on substrings in the name.
14:25 josephwb right
14:26 towodo wonder why they each get their own ncbi taxonomy record?  would be nice to know a bit about what's going on
14:28 towodo totally weird.  http://www.ncbi.nlm.nih.gov/taxonomy/?term=852484[uid]   is a taxonomy record with no sequences.  I thought that wasn't supposed to happen
14:28 towodo no metadata at all.
14:29 towodo there are 96587 BOLD: names in OTT 2.8
14:29 josephwb wha? that should happen (re: NCBI)
14:29 josephwb NCBI
14:30 towodo I always thought NCBI taxonomy only created a taxon record when there was a need to do so because of a genbank deposit.
14:30 towodo no sequence -> no taxon.
14:30 josephwb me too
14:31 towodo here we have an ad hoc taxon and no reference to a paper, or person, or sequence, or anything at all.
14:33 towodo never seen this before.
14:33 towodo the part after 'BOLD:' is not a genbank id
14:35 josephwb right, it is a BOLD ID (AFAIK)
14:38 mtholder joined #opentreeoflife
14:49 towodo so what is BOLD?  There is a bold Phidippus but it's not a BOLD Phidippus
14:50 josephwb barcode of life
14:50 josephwb http://www.boldsystems.org/index.php/TaxBrowser_Home
14:51 towodo ahh...
14:51 josephwb all COI sequences
14:52 josephwb part of the impetus is to sequence mulitple individuals of species
14:52 josephwb so, the 500 robin example is probably an underestimate
14:52 towodo so why aren't the sequences in genbank?
14:52 josephwb i have no idea
14:52 josephwb i think people are pissed that they are not
14:53 josephwb so, "Genus_sp._BOLD:XXX" is not useful
14:53 josephwb but I'd say that "Genus_species_BOLD:XXX" is also useless
14:54 towodo if the sequences are secrets then this doesn't like science.  didn't find any on web site
14:55 josephwb oh, it is a whole thing
14:55 towodo ahh… some of it's there… some data is 'published and released' and some isn't
14:55 josephwb i believe you need to make an account to see anything
14:56 towodo so this may be an embargo issue.
14:57 towodo the BOLD: ids are cluster ids
14:57 towodo 'Barcode Index Numbers(BIN): cluster barcode sequences to create OTUs that closely reflect species groupings '
14:57 towodo so it's like SILVA
14:57 josephwb so: crap
14:57 towodo http://www.boldsystems.org/index.php/Public_RecordView?processid=GBAN4447-13
14:58 towodo all but 5000 of the 90000 are 'unclassified' so will be filtered out by taxomachine
14:58 towodo oops treemachine
14:58 towodo not taxomachine
14:58 kcranstn 90000???
14:59 towodo yes, there are 90000 BOLD: 'taxa'
14:59 josephwb but all of the "Genus_sp._BOLD:XXX_otNNN" get through
14:59 towodo example?
14:59 josephwb i feel like this is new
15:00 towodo not new.  just as many in OTT 2.6.
15:00 josephwb Diptera sp. BOLD:AAN8686_OTT4513650
15:01 towodo that looks unclassified to me.
15:01 towodo 4513650 |       661378  |       Diptera sp. BOLD:AAN8686        |       species |       ncbi:972263     |               | \
15:01 towodo unclassified,sibling_higher
15:01 josephwb but i've never had this colon problem before (ha!)
15:02 josephwb a ":" in a name
15:02 josephwb let me re-build my taxonomy DB
15:02 josephwb i believe blackrim changed the import a bit (i.e. the change in 'classified')
15:03 towodo maybe some glitch in treemachine taxon filtering?  the 'unclassified' flag is new
15:03 towodo here's what it looks like in 2.6:
15:03 towodo 4513650 |       661378  |       Diptera sp. BOLD:AAN8686        |       species |       ncbi:972263     |               |       unclassified_inherited  |
15:03 towodo but I thought stephen added support for 'unclassified'.
15:04 towodo (vs. 'unclassified_inherited')
15:04 towodo are you running the latest treemachine?
15:04 josephwb right. i may have built my DB before that
15:04 josephwb but now i want to fix the ":" problem first
15:04 josephwb in case it is ever in a valid name
15:05 towodo in any case there were 5432 of these in OTT 2.6 not marked either 'unclassified' or 'unclassified_inherited'
15:06 kcranstn http://www.ncbi.nlm.nih.gov/nuccore/HM381713
15:07 towodo !!
15:08 towodo are these COI sequences likely to be used in phylogenetic studies?
15:19 mtholder joined #opentreeoflife
15:20 josephwb towodo: it is mainly intraspecific stuff, phylogeography, etc.
15:20 josephwb population genetics
15:20 josephwb could certainly be in a tree, but not a tree we are interested in
15:21 josephwb ok, fixed colon thingy
15:21 josephwb now off to rebuild DB
15:22 * mtholder just returned (after having forgotten to log out of IRC) and is catching up on the irc log...
15:22 towodo might want to check treemachine sources to make sure 'unclassified' gets filtered.
15:27 josephwb looks like it
15:29 mtholder side note: the issues w/ BOLD sequences were a big part of Rod's "dark taxa" posts a while back (see comment thread on http://iphylo.blogspot.com/2011/04/dark-taxa-genbank-in-post-taxonomic.html )
15:29 mtholder jimallman, I'll take a look at the reference issues...
15:30 jimallman thanks!
15:56 mtholder jimallman I'm pretty sure that the problem is that treemachine's getSyntheticSourceList is returning IDs as ##_## instead of ot_##_tree## (or, better yet {"study": "pg_#", "tree": "sometreeid"} )
15:57 mtholder oti is now using prefixed ids
15:57 mtholder so we are not finding any matches in the return from oti an the returned list from treemachine.
15:57 jimallman hm, last night i just got an empty list from getSyntheticSourceList… will try again
15:58 mtholder i think that treemachine db was being deployed yesterday (i suspect that is what made the machine sluggish for tree imports, too)
15:59 mtholder josephwb, blackrim: would it be tough to get the version of treemachine on ot10 to return namespace prefixed study IDs and treeIDs?
15:59 jimallman curl -X POST http://devapi.opentreeoflife.org/treemachine/v1/getSynthesisSourceList      â€¦ returns
16:00 jimallman [ "", "taxonomy" ]
16:00 mtholder oops. hold on I may be off base...
16:00 mtholder sorry I was pointing at ot12..
16:01 mtholder I'm getting the same from ot10. sorry for the noise.
16:01 blackrim jimallman: that is all that is in there now. wanted to update taxonomy so changed that DB to taxonomy. it will have other things in there shortly as we add sources. sorry about that
16:02 towodo well maybe we do need separate servers for curator's oti and for treemachine, given that the treemachine db will be reloaded daily.
16:03 blackrim josephwb: you may have already seen this but my output has the ott and underscores. not sure where the spaces came from
16:04 jimallman blackrim: no worries, just a minor mystery when testing the new Bibliographic References page.
16:04 josephwb tracking this down
16:11 jimallman Romina is asking about how to search for trees that contain one or more taxa. Do we have any working examples of this? I see it listed in the API todo list for hackathon, and notes that say it’s working now for tip taxa. Here’s my attempt. It seems to work, but obviously only searches for one taxon:
16:11 jimallman https://gist.github.com/jimallman/966bff09c95a25bce990
16:12 josephwb hmm, dendroscope doesn't like colons at all, even if they are in quotes
16:12 mtholder single quotes?
16:12 josephwb yup
16:12 josephwb tries double too
16:13 josephwb tried
16:14 mtholder fortunately, it is open source, so we can just... jk it is not open source
16:15 josephwb ha
16:15 mtholder joined #opentreeoflife
16:16 * mtholder is disconnecting for about 20 min.
16:16 towodo jimallman, the oti on ot10 has the taxomachine database included, so searches based on taxa inclusion should also work
16:16 towodo if docs don't say how then maybe we need to read the sources or consult cody
16:16 josephwb figtree handles it
16:16 josephwb but, of course, figtree cannot handle big trees
16:17 josephwb *crap*
16:17 jimallman towodo: yes, if there’s a multi-property (or multi-value) search method, i’m missing it.
16:17 josephwb trying to preserve taxon strings exactly
16:17 josephwb e.g. for searching
16:17 towodo oh.  sounds like it might be a lost cause.
16:18 towodo you could escape colons somehow
16:18 josephwb colon is the only character that doesn't work in quotes AFAIK
16:18 josephwb the quotes escape them
16:18 josephwb it is what they are for
16:19 towodo not standard apparently.
16:19 josephwb works for, say, '('
16:19 josephwb but it is!
16:19 josephwb must be a dendroscope bug
16:19 towodo is there a newick standard?  and does it matter, if an important program ignores it?
16:20 towodo https://en.wikipedia.org/wiki/Punycode  :')
16:20 josephwb it supports escaping all of the other illegal characters
16:20 blackrim can you edit a label in dendroscope to have a colon, then save it as newick and see if it escapes it somehow
16:21 josephwb http://evolution.genetics.washington.edu/phylip/newick_doc.html
16:21 josephwb will try
16:23 josephwb just replaves it with an underscore
16:23 blackrim so yeah, seems like it really doesn't like them
16:24 josephwb yup
16:24 josephwb originally we were stripping out all shitty characters
16:24 josephwb at the present we preserve them all
16:24 josephwb i can replace the ":" with "_" for a one-off
16:25 towodo I can't think of a reason why it has to be invertible.
16:25 towodo If you want to invert there's always the OTT id.
16:27 josephwb oh, newer dendroscope works
16:27 josephwb but has its own bugs
16:27 josephwb we use the older version for, well, everything
16:27 josephwb so i guess i will do the replacement
16:34 blackrim yeah, newer version is really crappy (selection doesn't work, seems buggier, etc.). at least neither is open source
16:37 towodo I want a nice lightweight tree browsing tool that doesn't do fancy graphics, maybe text-based
16:38 blackrim emacs
16:38 josephwb done with dendroscope fix
16:38 towodo keyboard macro + outline mode?  that would do
16:38 josephwb synth still broken for me
16:38 blackrim josephwb: what broke.
16:38 blackrim i am running one with some fungal trees and seems fine
16:41 mtholder joined #opentreeoflife
16:42 josephwb hmm
16:43 josephwb running into that weird Name_______NNNN thing with loading my tree
16:43 josephwb root name:Drosophilidae______34905
16:44 josephwb do you see that?
16:44 josephwb blackrim^
16:44 blackrim I am checking
16:45 blackrim no, mine seems to have name_ott#
16:46 blackrim are you doing load_synth_extract.run(.....
16:46 josephwb yes
16:47 josephwb seems like maybe my taxonomy is corrupted
16:47 josephwb so tree is not mapping
16:47 josephwb so i get taxonomy out
16:48 blackrim that is weird. i wonder what is going on there. are you doing load_taxonomy to load the taxonomy
16:48 josephwb no
16:48 josephwb i built it separately using inittax in treemachine
16:49 josephwb that must be it
16:49 blackrim probably pretty similar but if you check the load_taxonomy.py that is what I dowq
16:50 josephwb trying that now
17:06 blackrim mtholder: jimallman wondering where this study is http://www.reelab.net/phylografter/study/view/9 in curator or phylesystem
17:07 jimallman http://devtree.opentreeoflife.org/curator/study/view/pg_09
17:08 jimallman hm, searching for this (by DOI) in the curator’s study list shows two results. one from mtholder, another from bryan drew
17:09 jimallman bryan’s is http://devtree.opentreeoflife.org/curator/study/view/pg_1944
17:09 blackrim so you have to have a leading 0 for the single digit studies?
17:10 jimallman hm, that is sneaky but possible. i believe it has to do with shard subfolders using 2-digit numbers..? mtholder, is that correct?
17:15 blackrim jimallman: adding the 0 works.
17:15 mtholder sorry about that (the 2 digit thing). I can work on the aliasing so that either pg_9 or pg_09 will work.
17:16 blackrim it is fine, just needed to know for the plants file
17:18 mtholder there are 2 because there are 2 on phylografter. We'd have to check w/Bryan to see which one is correctly curated.
17:26 mtholder https://github.com/OpenTreeOfLife/phylesystem-1/issues/1 fwiw
17:28 josephwb blackrim: still have weird underscore thing happening
17:28 josephwb both in "root name:"
17:28 josephwb and
17:28 josephwb TEST
17:30 blackrim that is really weird. you are on master on both of those?
17:30 blackrim I am confused at how we are getting different things
17:32 josephwb yes, master
17:32 blackrim the file you are running is not pushed is it?
17:33 josephwb no
17:33 josephwb so, in treemachine when you initialize the taxonomy DB, you have to give it a name e.g. ott2.8
17:34 josephwb i don't think the gcmdr scripts do that
17:34 blackrim why do you have to give it a name?
17:34 josephwb when i query for the taxonomy version from the gcmdr-initialized DB i get the name of the synonyms file
17:35 josephwb so if you have multiple DBs lying around, you can look up which version of ott was used to build it
17:35 josephwb e.g. trying to add a tree mapped with 2.8 into a DB built from 2.6 would not be good
17:36 blackrim ok, but the inittax doesn't take that as arguments does it (or did this change at some point)
17:36 josephwb yes it does
17:36 josephwb synonym file is optional
17:37 josephwb but taxonomy name is not
17:37 josephwb if you don't think this is useful i can excise it
17:38 blackrim yeah, i am not sure it is useful. i can see how the information is useful to serve but I don't thikn we have the version that the taxa were mapped with
17:39 blackrim it is fine to have the name though, I can change that in gcmdr (but it also needs to be changed in the doc of treemachine)
17:39 blackrim java -jar target/treemachine-0.0.1-SNAPSHOT-jar-with-dependencies.jar
17:39 blackrim just says
17:39 blackrim inittax <filename> (optional:synonymfilename) <graphdbfolder>
17:39 josephwb that's not right
17:39 blackrim if you do -help you get the longer one
17:40 josephwb oh, i guess i didn't fix the shorter one.
17:40 blackrim it is fine, I can change it in gcmdr
17:40 josephwb https://github.com/OpenTreeOfLife/treemachine/commit/ab30937c835cb577218d9b1e82a5f4d1bf388538
17:40 josephwb up to you
17:41 josephwb should study files have an annotation of what version of ott was used to map the taxa?
17:41 josephwb seems the curator app could do this.
17:41 blackrim the problem with that is that they will often be behind because you dont' update the mapping automatically with updated taxonomy
17:41 josephwb right
17:41 blackrim so you could have that it was 2.6 and no one edited it since 2.8 but ott ids persist so not a problem
17:42 blackrim of course, we don't really use synonyms much right now so not a big deal (and likely not causing your error)
17:42 josephwb right, it was the other way around that might be a problem.
17:43 blackrim ok, gcmdr should be right
17:43 josephwb i.e. if mapped in 2.8, might have taxa that were not present in 2.6
17:44 josephwb so, all these DB building we've been doing has taken in one of the arguments as the taxonomy version. probably means synonyms was lost from the db.
17:45 josephwb not a big deal, as we have everything mapped
17:45 josephwb so, yeah, my problem: no idea
17:49 josephwb blackrim: i am doing study: pg_1940_3943
17:49 josephwb synthottolid=661378 (Diptera)
17:50 josephwb this is the Drosophilidae tree
18:19 jimallman josephwb: regarding saving the mapping-OTT version in curation app, it’s an interesting idea. but we’d need to save this for every mapped OTU, since it’s possible to revisit and continue mapping OTUs at any time. Or to unmap and remap a single OTU.
18:20 josephwb right
18:20 jimallman we’ve talked in the past about automated “sweeps” of study Nexson to update mapping to the latest. but as far as i known, no final decision was reached.
18:20 mtholder i thought the IDs were stable. why do we need to know what version of OTT was used?
18:21 josephwb when a taxon is moved, does it keep the same ottid?
18:21 mtholder towodo^
18:22 josephwb e.g. a genus is moved from one family to another
18:22 josephwb if so, knowing the ottversion might be useful
18:22 * towodo in a meeting
18:23 mtholder I guess if we have some new synonym discovered in a new version of OTT (and that synonym was a taxon name in a previous OTT), we should go through the corpus of mapped studies and flag anything mapped to that synonym (or delete the mapping if we were being aggressive)
18:24 josephwb i just thought this might be a nice piece of info to have
18:24 josephwb when the taxonomy was changing quickly, it was hard to keep track of DB origins
18:24 josephwb and changes to taxonomy would lead to greatly different trees
18:25 mtholder It would be nice to be able to recreate the TNRS results that the curator would have seen, but that sounds expensive.
18:25 jimallman left #opentreeoflife
18:26 towodo the answer is yes, the taxon keeps the id
18:26 towodo there are exceptions in pathological situations
18:27 josephwb so taxonomies can contain the exact same taxa (ottids), but configured differently
18:28 josephwb i am not pushing for anything here. i just thought we'd want to record this stuff, since reproducability sometimes comes up
18:29 towodo I think cody put a draft paper on google drive somewhere
18:32 josephwb i seem to be the only one who thought this might be a good idea, and i am out to lunch, so forget it
18:33 mtholder I think it would be nice, but it seems like we'd need a whole array of taxomachines serving up each of the legacy versions of TNRS if we wanted to be able to peek back at what the curator saw.
18:33 mtholder sounds tough
18:33 mtholder and I'm not sure how often we'd use it.
18:34 mtholder If we keep an accurate log of when we redeploy new versions of OTT to taxomachine (and when we change where the curator app looks for TNRS), then we could infer the OTT versions from the commit log.
18:35 mtholder perhaps, we could accomplish this with an annotation in every commit that changed OTU mapping. I'd like to avoid adding an "ottVersion" to every mapped OTU, but there may be a lightweight way of storing this info.
18:37 josephwb i guess i was thinking a global value for the whole study
18:38 josephwb when ott updates, process the study to see how/if it conforms
19:02 towodo Did taxomachine get reloaded when we went to 2.8?
19:03 mtholder we should be able to check: https://github.com/OpenTreeOfLife/taxomachine/issues/37
19:07 mtholder though neither the instance of devapi on devapi nor the one on api return that taxonomy issue.
19:07 mtholder s/issue/object/
19:08 mtholder curl -X POST http://api.opentreeoflife.org/taxomachine/v1/contextQueryForNames -H "content-type:application/json" -d '{"queryString":"Trillium decipiens"}'
19:08 mtholder curl -X POST http://devapi.opentreeoflife.org/taxomachine/v1/contextQueryForNames -H "content-type:application/json" -d '{"queryString":"Trillium decipiens"}'
19:30 towodo So curation is still stuck at OTT 2.6.
19:30 josephwb good to know
19:30 towodo or is it.  I'll look at the downloads dir
19:32 towodo taxomachine.db.tgz was writtn on June 10.  that was when I copied it from ot9.  so yes, we're still at OTT 2.6.
19:33 josephwb i had suspicions
19:33 towodo but it doesn't matter in the instance Romina cites.
19:33 josephwb re: the mapping of some taxa but not other that we spoke of the other day
19:33 mtholder the curator is running on ot11 (devtree), right? OPENTREE_NEO4J_HOST=devapi.opentreeoflife.org in that config
19:33 mtholder nm "from ot9"
19:34 towodo sorry, I was talking about taxomachine.  taxomachine's db on ot10 was copied from ot9.
19:34 mtholder got it.
19:35 towodo ott2.8 is new, taxomachine on ot10 hasn't caught up with it.  stephen was going to take it out for a spin first
19:38 josephwb blackrim: got things fixed with my synth
19:39 josephwb seems like an unnecessary regex replacement was necessary after all
20:16 jimallman joined #opentreeoflife
20:19 jimallman joined #opentreeoflife
20:28 mtholder joined #opentreeoflife
20:36 17SAAC7EF joined #opentreeoflife
21:31 josephwb joined #opentreeoflife
21:42 josephwb joined #opentreeoflife
21:42 mtholder joined #opentreeoflife
22:51 mtholder joined #opentreeoflife
23:15 josephwb joined #opentreeoflife

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