Wikidata:Property proposal/parent form of an active substance

‎parent form of an active substance edit

Originally proposed at Wikidata:Property proposal/Natural science

Done: parent form of an active substance (P12099) (Talk and documentation)

Description	parent form of a chemical entity used as an active substance in the form of a salt
Data type	Item
Domain	chemical entity
Example 1	(−)-pseudoephedrine hydrochloride (Q83093654)this property(−)-pseudoephedrine (Q27288764)
Example 2	levamfetamine sulfate (Q27286830)this property(R)-amphetamine (Q2506823)
Example 3	eletriptan hydrobromide (Q27130865)this propertyeletriptan (Q415032)
Single-value constraint	yes

Motywacja edit

In many cases active substances are used in the form of salts. Currently we do not have any way to link between the active substance and the chemical form it is used in pharmaceutical formulations. We have a way to link between an active substance and a pharmaceutical product with a has active ingredient (P3781)/active ingredient in (P3780) (however, this is not consistently applied to items).

This could be also done in the reverse manner, i.e. parent substancethis propertysalt, but the proposed option seems to be consistent with other properties like hydrated form of (P4770). Wostr (talk) 15:36, 9 October 2023 (UTC)[reply]

Discussion edit

Remind me: what is the standard vocabulary for describing the relationship between a chemical and its salt? Bluerasberry (talk) 15:41, 9 October 2023 (UTC)[reply]

In WD we don't have any mean to do it right now (and I'm not sure we should have for all chemical entities). It is possible to use different approaches: in ChEBI there is a 'has_functional_parent' relationship for some entities, but there is no direct way to do what is proposed here. The ChEBI would link e.g. D-methamphetamine (Q191924) to its conjugated acid via methamphetamine(1+) (Q27225520)is_conjugated_acid_ofD-methamphetamine (Q191924), then the salt form would be linked to the conjugated acid via 'has_part' methamphetamine hydrochloride (Q27116464)has_partmethamphetamine(1+) (Q27225520). This way we don't have any direct relationship, also we can't even easily tell when such a relation links an active substance to its salt form. And in WD we can't use D-methamphetamine (Q191924)part_ofmethamphetamine hydrochloride (Q27116464) as it's not true (the salt is composed of a protonated form of an active substance for which we should have different item, just like in the example from ChEBI). Some medicine-related databases uses 'subclass_of" to link between an active substance and its salt, but this approach wouldn't be correct from chemical POV. UNII uses 'PARENT->SALT/SOLVATE' relation. Wostr (talk) 16:00, 9 October 2023 (UTC)[reply]

BTW, we also have https://www.wikidata.org/wiki/Q106209873#P279 but I heard usage of "of" is frowned upon. --SCIdude (talk) 11:23, 11 October 2023 (UTC)[reply]

Very ambiguous qualifier, in many languages there is no equivalent word for the English 'of', so the label describing the scope of this qualifier is a kind of workaround (and in fact, this is why the scope between languages may differ). Also, it does not allow any simple actions, e.g. via sparql; some derivatives may be complex enough to not be considered or used as the form of an active substance. I would opt for not using this qualifier at all. Wostr (talk) 11:55, 12 October 2023 (UTC)[reply]

Support Useful property --Ameisenigel (talk) 20:51, 9 October 2023 (UTC)[reply]
Support In principle, has part/is part of can be used for this, but the suggested property is pretty standard terminology in several databases. I think we should do this. --Egon Willighagen (talk) 03:57, 10 October 2023 (UTC)[reply]
Support. --Csisc (talk) 21:33, 10 October 2023 (UTC)[reply]
@Wostr, Ameisenigel, Egon Willighagen, Csisc: Done --Tinker Bell ★ ♥ 19:26, 18 October 2023 (UTC)[reply]