kpmp / ktao Goto Github PK

Jonathan Barasch should weigh in on the up to date mappings of the sequencing derived ontology mappings. I was reviewing the 5-2018 release of KTAO vs the recent Science paper he is on, and they have additional abbreviations and proposed new cells we should probably include or map.

provisional mappings:
| 1 | Endo: endothelial, vascular and descending loop of Henle

| 1-1 | Endo: endothelial
http://purl.obolibrary.org/obo/CL_0000115
-- many different subtypes in the ontology

| 1-2 | Peri: vascular
?? pericyte http://purl.obolibrary.org/obo/CL_1001318
confirm mesangial and interstitial pericytes included in this class

| 1-3 | DLH: descending loop of Henle
http://purl.obolibrary.org/obo/CL_1001021

| 2 | Podo: podocyte
http://purl.obolibrary.org/obo/CL_1000451
http://purl.obolibrary.org/obo/CL_0000653

| 3 | PT: proximal tubule
http://purl.obolibrary.org/obo/CL_1000839

| 3-1 | S1 PCT: S1 segment of proximal tubule, proximal convoluted tubule
http://purl.obolibrary.org/obo/UBERON_0001287
http://purl.obolibrary.org/obo/UBERON_0005124

| 3-2 | S2: S2 segment of proximal tubule
http://purl.obolibrary.org/obo/UBERON_0004197
--NEED to add this

| 3-3 | S3 PST: S3 segment of proximal tubule, proximal straight tubule
http://purl.obolibrary.org/obo/UBERON_0001290

| 4 | LOH: loop of henle

http://purl.obolibrary.org/obo/UBERON_0001288
http://purl.obolibrary.org/obo/CL_1001109
http://purl.obolibrary.org/obo/CL_1001108
http://purl.obolibrary.org/obo/CL_1001107
http://purl.obolibrary.org/obo/CL_1001106

| 5 | DCT: distal convoluted tubule
http://purl.obolibrary.org/obo/UBERON_0005118
http://purl.obolibrary.org/obo/UBERON_0005120

| 6 | CD-PC: collecting duct principal cell
http://purl.obolibrary.org/obo/CL_1000550

| 7 | CD-IC: CD intercalated cell|collecting duct intercalated cell
http://purl.obolibrary.org/obo/CL_0005010
| 7-1 | A-IC: alpha intercalated cells|α-intercalated cells
http://purl.obolibrary.org/obo/CL_0005011
| 7-2 | B-IC: beta intercalated cells
http://purl.obolibrary.org/obo/CL_0002201
--Need to add

I don't see the remaining cells in the ontology
--Do we need to add new entries or links?

| 8 | CD-Trans: CD transient cell

| 9 | novel cell type 1
?? some type of proliferative or stem cell like PT cells
-- Do we need to add placeholder pending final publication describing?

| 10 | Fib: fibroblast
http://purl.obolibrary.org/obo/CL_0000057
| 11 | Macro: macrophage
http://purl.obolibrary.org/obo/CL_0000235
| 12 | Neutro: neutrophil
http://purl.obolibrary.org/obo/CL_0000096
| 13 | B lymphocyte
http://purl.obolibrary.org/obo/CL_0000236
| 14 | T lymphocyte
http://purl.obolibrary.org/obo/CL_0000084
| 15 | NK: natural killer cell.
http://purl.obolibrary.org/obo/CL_0000623

| 16 | novel cell type 2
-- Do we need to add placeholder pending final publication describing?

Summary proposed entities to add from NCBO
http://purl.obolibrary.org/obo/CL_0002201
http://purl.obolibrary.org/obo/UBERON_0004197

http://purl.obolibrary.org/obo/CL_0000057
http://purl.obolibrary.org/obo/CL_0000235
http://purl.obolibrary.org/obo/CL_0000096
http://purl.obolibrary.org/obo/CL_0000236
http://purl.obolibrary.org/obo/CL_0000084
http://purl.obolibrary.org/obo/CL_0000623

Summary proposed new entities
Collecting duct transient cell
novel cell type 1
novel cell type 2

Running HermiT on ktao-merged also throws the following error:

Input ontology contains 5 triple(s) that could not be parsed:
 - <http://purl.obolibrary.org/obo/OPMI_0000317> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> _:genid2147502689.
 - <http://purl.obolibrary.org/obo/OPMI_0000315> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> _:genid2147502674.
 - <http://purl.obolibrary.org/obo/OPMI_0000316> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> _:genid2147502684.
 - <http://purl.obolibrary.org/obo/OBI_0000759> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> _:genid2147502673.
 - <http://purl.obolibrary.org/obo/OPMI_0000315> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> _:genid2147502679.

If you open the ontology in Protege and navigate to each of these individuals, you will find object property assertions in the type assertion box. Simply making those type assertions into object property assertions should fix this.

If you run HermiT on ktao-merged.owl, it throws the following error: The given property hierarchy is not regular. There is a cyclic dependency involving property <http://purl.obolibrary.org/obo/BFO_0000050>.

Basically, OWL 2 DL places global restrictions on object properties to ensure that OWL 2 DL is decidable. One of these restrictions is that you are not allowed to use a transitive property, like part-of (BFO:0000050), in a property chain axiom, like in the example below:

<owl:propertyChainAxiom rdf:parseType="Collection"> 
        <rdf:Description rdf:about="http://purl.obolibrary.org/obo/BFO_0000050"/>  
        <rdf:Description rdf:about="http://purl.obolibrary.org/obo/BFO_0000063"/>  
</owl:propertyChainAxiom>

There seems to be quite a few more of these axioms, other than this one, in the ontology.

This is an issue out of a discussion on April 3, 2018 between Jonas M. Carson and Oliver He:

We need to handle KPMP specimen and their derivatives using ontology, e.g., as Dr. Sanjay Jain from UW mentioned:
"Kidney-frozen OCT block Dry ice- sections-RNA/protein/dna/metabolite/fixed-stains-HE/immuno
Kidney-frozen LN2- frozen sections/RNA/protein/dna/metabolite
Kidney-frozen RNAlater- RNA
Kidney-fixed paraffin block-sections-stain-HE/PAS/JMS/Trichrome/immuno
Kidney-fixed cryopreserved frozen OCT block-sections- RNA/fixed-stains-HE/immuno stain-HE/PAS/JMS/Trichrome/immuno
Kidney-dissociated suspension- cells/nuclei-RNA/protein

Once a specimen ID is registered with basic metadata, the user can choose the samples associated with that and add derivatives (how many different tissue pieces made from the parent tissue, what was further derived from them such block, RNA, protein, metabolites etc) and how many aliquots (example-sections or slides, number of RNA aliquots) are made so for subsequent distributions these can be tracked and inventory kept."

For this use case, we can seperate the examples based on categories:

• Freezing methods:
  o OCT block dry ice
  o LN2- frozen
  o RNAlater
• Fixation:
  o paraffin block
  o cryopreserved
• Section:
  o RNA
  o Protein
  o DNA
  o Metabolite
• Staining
  o HE
  o Immune
  o PAS
  o JMS
  o Trichrome
• Dissociated suspension
  o Cell
  o Nuclei-RNA
  o Protein
• Number of aliquots

We may also use the Ontology for BioBanking: http://www.ontobee.org/ontology/OBIB. However, it appears OBIB is not sufficient to handle this issue. We may contact and collaborate with them. Also, such a study may become part of the KTAO. Any suggestions and comments?

In the future we plan on reaching out to developers of ontologies that import our classes to get their input prior to making such changes, but we have replaced this class with another one with the same label: 'role in human social processes' (OMRSE:00002072). The only change to the class was to the subClassOf axiom, which we changed from "inheres-in only human or organization" to "inhere-in only 'material entity'."