Re: [gpc-informatics] #158: usable view of LOINC lab terms

[GPC Informatics]

#158: usable view of LOINC lab terms
-+
 Reporter:  rwaitman |   Owner:  budh0007
 Type:  enhancement  |  Status:  assigned
 Priority:  major|   Milestone:  data-domains2
Component:  data-stds|  Resolution:
 Keywords:   |  Blocked By:
 Blocking:  68   |
-+

Comment (by nateapathy):

 Lucky for us, Lori already did some of this in January. Unfortunately, the
 LOINC file is ~460MB so I won't be uploading/attaching it here. Here's the
 link to the site specifically for the Bioportal-to-i2b2 translations:
 http://i2b2.bioontology.org/

 And here's the link to the Bioportal API tool:
 https://community.i2b2.org/wiki/display/NCBO/NCBO+Ontology+Tools

--
Ticket URL: 

gpc-informatics 
Greater Plains Network - Informatics
___
Gpc-dev mailing list
Gpc-dev@listserv.kumc.edu
http://listserv.kumc.edu/mailman/listinfo/gpc-dev

Re: [gpc-informatics] #158: usable view of LOINC lab terms

[GPC Informatics]

#158: usable view of LOINC lab terms
-+
 Reporter:  rwaitman |   Owner:  budh0007
 Type:  enhancement  |  Status:  assigned
 Priority:  major|   Milestone:  data-domains2
Component:  data-stds|  Resolution:
 Keywords:   |  Blocked By:
 Blocking:  68   |
-+

Comment (by nateapathy):

 Right - we don't keep that top level of the hierarchy, only below the
 LOINCPARTS section is used.

 And the synonym question is a good one; that answer goes beyond my
 familiarity with Bioportal. Why that would be picked as a "''preferred''
 name" would be even more interesting to know. I'll see if we can pull down
 a raw csv of the export using Lori's API tool and post to this ticket for
 combing through how we might swap those names around in the i2b2
 ontology/metadata tables.

--
Ticket URL: 

gpc-informatics 
Greater Plains Network - Informatics
___
Gpc-dev mailing list
Gpc-dev@listserv.kumc.edu
http://listserv.kumc.edu/mailman/listinfo/gpc-dev

Re: [gpc-informatics] #158: usable view of LOINC lab terms

[GPC Informatics]

#158: usable view of LOINC lab terms
-+
 Reporter:  rwaitman |   Owner:  budh0007
 Type:  enhancement  |  Status:  assigned
 Priority:  major|   Milestone:  data-domains2
Component:  data-stds|  Resolution:
 Keywords:   |  Blocked By:
 Blocking:  68   |
-+

Comment (by dconnolly):

 Yes, that's the sort of think I was looking for. But...

 I don't know many clinical researchers who could handle the first choice
 it presents:

  - LOINCCLASSTYPES
  - LOINCPARTS

 Below LOINCPARTS looks OK. And it seems to match CMH's labs on babel. It's
 somewhat similar to the "GPC: Laboratory measurements" area, but where CMH
 has Chemistry (LP31388-9), GPC has Chemistry non challenge tests (7786-9)
 with over 500 children. So yes, the Bioportal ontology looks like a
 worthwhile path to pursue.

 The commonly used potassium code (2823-3) wasn't too hard to find, but the
 label is kinda cryptic: "Potassium:SCnc:Pt:Ser/Plas:Qn". I wonder why they
 didn't choose the synonym "Potassium [Moles/volume] in Serum or Plasma".
 And I wonder if there's a straightforward way to get nicer labels like
 that throughout.

--
Ticket URL: 

gpc-informatics 
Greater Plains Network - Informatics
___
Gpc-dev mailing list
Gpc-dev@listserv.kumc.edu
http://listserv.kumc.edu/mailman/listinfo/gpc-dev

Re: [gpc-informatics] #158: usable view of LOINC lab terms

[GPC Informatics]

#158: usable view of LOINC lab terms
-+
 Reporter:  rwaitman |   Owner:  budh0007
 Type:  enhancement  |  Status:  assigned
 Priority:  major|   Milestone:  data-domains2
Component:  data-stds|  Resolution:
 Keywords:   |  Blocked By:
 Blocking:  68   |
-+

Comment (by nateapathy):

 Try taking a look
 
[http://bioportal.bioontology.org/ontologies/LOINC/?p=classes&conceptid=http%3A%2F%2Fpurl.bioontology.org%2Fontology%2FLNC%2FMTHU000999
 here] (by clicking on the "2014AA" you can get a navigable/browsable
 hierarchy) and expand out the LOINCPARTS section to browse. Is that what
 you're looking for?

--
Ticket URL: 

gpc-informatics 
Greater Plains Network - Informatics
___
Gpc-dev mailing list
Gpc-dev@listserv.kumc.edu
http://listserv.kumc.edu/mailman/listinfo/gpc-dev

Re: [gpc-informatics] #158: usable view of LOINC lab terms

[GPC Informatics]

#158: usable view of LOINC lab terms
-+
 Reporter:  rwaitman |   Owner:  budh0007
 Type:  enhancement  |  Status:  assigned
 Priority:  major|   Milestone:  data-domains2
Component:  data-stds|  Resolution:
 Keywords:   |  Blocked By:
 Blocking:  68   |
-+

Comment (by dconnolly):

 Replying to [comment:4 nateapathy]:
 > ... Have other sites used the Bioportal ontology? Or am I missing
 documentation on that investigation for GPC terms? And if not, would that
 be a worthwhile path to pursue?

 I hadn't looked at it closely until now, but "Maximum number of children:
 10329" won't work well in the i2b2 Navigate Terms UI. There are also "No
 views available" so it's not convenient to browse. There is a "Create new
 view" button. Is it straightforward to make a usable view?

--
Ticket URL: 

gpc-informatics 
Greater Plains Network - Informatics
___
Gpc-dev mailing list
Gpc-dev@listserv.kumc.edu
http://listserv.kumc.edu/mailman/listinfo/gpc-dev

Re: [gpc-informatics] #158: usable view of LOINC lab terms

[GPC Informatics]

#158: usable view of LOINC lab terms
-+
 Reporter:  rwaitman |   Owner:  budh0007
 Type:  enhancement  |  Status:  assigned
 Priority:  major|   Milestone:  data-domains2
Component:  data-stds|  Resolution:
 Keywords:   |  Blocked By:
 Blocking:  68   |
-+

Comment (by dconnolly):

 It would be better if we had a (more) objective definition of usable, but
 part of the "reporter" role is judging when things are done. So we can
 expect Russ to judge proposals. My understanding is that the goal is
 something like: it's usable when Russ doesn't feel embarrassed to ask
 researchers to use it.

 Another approach might be: ask a handful of clinical researchers to find
 potassium and a couple others and see how well they do. If they do well, I
 expect that would be compelling evidence that this is done.

--
Ticket URL: 

gpc-informatics 
Greater Plains Network - Informatics
___
Gpc-dev mailing list
Gpc-dev@listserv.kumc.edu
http://listserv.kumc.edu/mailman/listinfo/gpc-dev

Re: [gpc-informatics] #158: usable view of LOINC lab terms

[GPC Informatics]

#158: usable view of LOINC lab terms
-+
 Reporter:  rwaitman |   Owner:  budh0007
 Type:  enhancement  |  Status:  assigned
 Priority:  major|   Milestone:  data-domains2
Component:  data-stds|  Resolution:
 Keywords:   |  Blocked By:
 Blocking:  68   |
-+

Comment (by nateapathy):

 In reviewing the lab lead responsibilities, it appears CMH/Cerner is
 responsible for driving this standardization. What is the status of the
 UMN student who was working on a project around this?

 Also, having looked through the documentation on tickets and emails that I
 can find regarding this, I don't see any references to
 http://bioportal.bioontology.org/ontologies/LOINC, which is where we
 derived our LOINC ontology from. They keep versions very up-to-date
 (latest maintained version is from 5/28/14), and the Bioportal site has an
 API that Lori Phillips at Partners has built an "i2b2 ontology builder"
 application on top of. This avoids the issue mentioned above re:
 downloading an Access database from REMLA every time we need to update the
 hierarchy. Have other sites used the Bioportal ontology? Or am I missing
 documentation on that investigation for GPC terms? And if not, would that
 be a worthwhile path to pursue?

 My final question is around the definition of "sufficiently usable" in the
 description of this ticket, which seems opaque to me unless I am missing a
 definition that the group decided upon earlier in the LOINC ontology
 process.

--
Ticket URL: 

gpc-informatics 
Greater Plains Network - Informatics
___
Gpc-dev mailing list
Gpc-dev@listserv.kumc.edu
http://listserv.kumc.edu/mailman/listinfo/gpc-dev