Wikidata:Property proposal/pathway annotation

pathway annotation

Originally proposed at Wikidata:Property proposal/Natural science

Withdrawn

Description	MISSING
Represents	annotation (Q857525)
Data type	Item
Domain	biological pathway (Q4915012)
Allowed values	Wikidata item (Q16222597)
Example 1	Overview of leukocyte-intrinsic Hippo pathway functions (Q66104607) → cancer (Q12078)
Example 2	Overview of leukocyte-intrinsic Hippo pathway functions (Q66104607) → hippo signaling (Q14914369)
Example 3	Overview of leukocyte-intrinsic Hippo pathway functions (Q66104607) → immune response-regulating signaling pathway (Q21108327)
Example 4	T-cell antigen receptor (TCR) pathway during Staphylococcus aureus infection (Q30225430) → immune response-regulating signaling pathway (Q21108327)
Planned use	to add statements to hundreds of existing biological pathway items

Motivation

Working on annotating biological pathway items. Linking these to annotations, like disease and pathway ontology terms, will enhance their value in Wikidata. --AlexanderPico (talk) 00:03, 9 October 2019 (UTC)[reply]

Discussion

PubMed (near bottom of page) and Identifiers.org use the term "keywords" for basically this purpose.

"keywords" are indeed commonly used in scientific publishing. They are very generic, however, and not synonymous or as clearly defined as "pathway annotation". The "main subject" property includes "keywords" as one of over a dozen aliases. This property is arguably even more vague than "keywords". So, the proposal stands for a new property with more clearly defined use. --AlexanderPico (talk) 17:32, 10 October 2019 (UTC)[reply]

Context for background/future work and reasons why they are not appropriate for this proposed use:

Prader-Willi and Angelman syndrome (Q44013803) → main subject → Prader–Willi syndrome (Q594013) + Angelman syndrome (Q535364) [Too vague]
Purine metabolism and related disorders (Q54989372) → medical condition → adenylosuccinase lyase deficiency (Q4682317) [Too specific]
Overview of leukocyte-intrinsic Hippo pathway functions (Q66104607) → related category → Category:Cancer (Q9699097) [Only for categories]
Prader-Willi and Angelman syndrome (Q44013803) → genre → disease pathway (Q28864297) [Breaks definition of genre, which is nicely scoped by current aliases]

WikiProject Molecular biology has more than 50 participants and couldn't be pinged. Please post on the WikiProject's talk page instead.

Oppose See discussion below. AlexanderPico (talk) 00:15, 9 October 2019 (UTC)[reply]
Support Andrew Su (talk) 04:43, 9 October 2019 (UTC)[reply]
Support Andrawaag (talk) 17:28, 9 October 2019 (UTC)[reply]
Support. YULdigitalpreservation (talk) 17:30, 9 October 2019 (UTC)[reply]
Wait Ariutta (talk) 00:45, 10 October 2019 (UTC)[reply]
Support David (talk) 14:21, 10 October 2019 (UTC)[reply]
Oppose Excellent point, but it is a very broad property, meaning "some kind of relation". For the examples in the proposal, a modification of main subject to accomodate biological pathways would be enough. A biological pathway is a set of inter-connected reactions and interactions whose delineation and scope are used as a model". The set assembled by the community can be seen as a piece of work, subjected to curation and changes, which has "main subjects". TiagoLubiana (talk) 20:37, 10 October 2019 (UTC)[reply]

Right. "main subject" is certainly broad enough to encompass our intended use here (like the examples in the proposal). But since we have a very specific use in mind, we thought it would be better to mint a more specific property. It's not just the specificity of subjects (bioloigcal pathways), but also the specificity of objects (ontology-backed annotations), like diseases, cell types and processes. If I could capture more of this specificity in the proposal, would you be persuaded? Or is your argument that since an existing property *could* work (frankly, because it is so vague) we should just use it? --AlexanderPico (talk) 00:12, 11 October 2019 (UTC)[reply]

The ontologies include disease, cell type and pathway type. Option #1 is to use pathway_annotation for all, e.g.: pathway → pathway_annotation → cancer + pancreatic_ductal_cell + cancer_pathway. Option #2 would just replace pathway_annotation with main_subject. Option #3 is to use a different property for each ontology, e.g., something like pathway → medical_condition → cancer; pathway → anatomical_location → pancreatic_ductal_cell; pathway → main_subject → cancer_pathway. --Ariutta (talk) 23:56, 11 October 2019 (UTC)[reply]

Ooh, option #3 sounds like a "baby bear" solution: not too vague, not too specific, but rather just right! Should I retract this proposal? How do I formally do that? --AlexanderPico (talk) 02:21, 12 October 2019 (UTC)[reply]

If you want to withdraw it you can edit the for above to say "status = withdrawn". ChristianKl ❪✉❫ 16:27, 14 October 2019 (UTC)[reply]

This doesn't even have a property description at the moment. Oppose given that this isn't specific enough and there's no attempt made at explaining why we would want to be vague here. How would this an improvement over medical condition (P1050) that's currently used to describe the relationship between Overview of leukocyte-intrinsic Hippo pathway functions (Q66104607) and cancer (Q12078)? ChristianKl ❪✉❫ 16:25, 14 October 2019 (UTC)[reply]

Agreed. See discussion right above this one. How do I formally close or withdraw this proposal? I'll try removing it from the list. --AlexanderPico (talk) 17:15, 14 October 2019 (UTC)[reply]

I just noticed that Overview of leukocyte-intrinsic Hippo pathway functions (Q66104607) seems to be mixed between being an item about a paper and being an item about a pathway. It likely shouldn't mix both given that they are different ontological entities. ChristianKl ❪✉❫ 16:31, 14 October 2019 (UTC)[reply]

Overview of leukocyte-intrinsic Hippo pathway functions (Q66104607) is about a pathway model. It cites a paper that was used as a reference to construct the model. The paper is part of the pathway's bibliography . --AlexanderPico (talk) 17:15, 14 October 2019 (UTC)[reply]

@AlexanderPico: A pathway doesn't cite a paper or has authors. It also shouldn't be named 'Overview of'. An item about the paper could link to the pathway item with main subject (P921). ChristianKl ❪✉❫ 12:41, 23 October 2019 (UTC)[reply]