LOINC-inspired clinical assay template for discussion on Monday 13th

[turbomam]

@handemcginty and I will be discussing clinical assay modelling on Monday April 13th. I'll specifically be discussing the work @cstoeckert and I have been doing, with lots of good input form the community already.

I'll post a link to the ROBOT template before Monday morning.

[turbomam]

Here's our latest clinical assay template:

https://github.com/PennTURBO/loinc_in_obo/blob/master/lobo_template_20200413am.csv

[turbomam]

@cstoeckert, @jamesaoverton and I have been working on these tasks:

• place all of these new assays as subClassesOf a new "assay on specimen from organism" class (in the TURBO ontology for now)
  • these assays can be applied to any specimen form any organism, regardless of research vs clinical objectives
  • how to handle new assays that don't take a specimen as the input, like an X-ray?
  • @cstoeckert and I will still need to create some new specimen types and modify our code to make the connections between anatomical entities and specimens
• compare IRI substrings or pre- and post-reasoned ontologies to detect overlap between existing OBI assays and new assays from this module
  • previous plans to check for owl:equivalentClasses won't work because OBI assays are defined as subClasses
  • in any case, detecting equivalentClasses with OBI requires several run time flags.

[linikujp]

Please check this CIDO-ontology/cido#7 for COVID-19 related specimen.
And I am working on things with FDA EUA PCR tests and mapping to LOINC. We should work together on this. Thanks!

[turbomam]

Hi @linikujp !

I got your email and will reply today.

[linikujp]

Maybe useful for you: https://www.hhs.gov/sites/default/files/covid-19-laboratory-data-reporting-guidance.pdf

[cstoeckert]

We are not generating assays based on LOINC due to licensing concerns. Instead we are generating them based on fields in Penn EHR records for lab tests and results.