User talk:DeSl

From Wikidata
Jump to navigation Jump to search

Can you please add some information to this page? Check my profile page to see how you can add your ORCID here. --Egon Willighagen (talk) 15:57, 11 July 2017 (UTC)Reply[reply]

"Instance of" or "Subclass of"[edit]

Denise, I think for compound classes you should use subclass of (P279) instead of instance of (P31), e.g. for pentose phosphate (Q32904830) where you use instance of (P31) now. --Egon Willighagen (talk) 13:33, 12 July 2017 (UTC)Reply[reply]

@Egon_Willighagen: okay, I looked at the examples. So instance of means: this is an example. With subclass, you mean: an example which has a hierarchy?
@Egon_Willighagen: thank you for explaining the talk page to me, let's see if I got it ;)

racemic/isomeric mixtures[edit]

Thanks for tackling that issue! Regarding the edits you made for D-Cysteine (Q16633812), my suggestion is to add part of (P361) to all pure compounds which are part of a racemic/isomeric mixture and use has part(s) (P527) to link from the mixture to the single compounds. The indiviual stereisomers can be linked with the property stereoisomer of (P3364). Thanks, Sebotic (talk) 21:25, 1 December 2017 (UTC)Reply[reply]

@Sebotic: : Thank you for this tip. What I saw up to now, is that other users use the "part of/has part" for their atomic composition; I didn't know that I could also use this for mixtures, but I will change this accordingly (have been working on proteinogenic amino acids and their isomers last week in detail).

Please check before creating new items for chemical[edit]

Hi, please can you check if a chemicla already exists before creating new items ? You created L-2-aminoadipic acid (Q44075060) but this chemical already exists as α-aminoadipic acid (Q2823204). So I converted your item into one of the two isomers of α-aminoadipic acid (Q2823204).

The simple way to check is to add the name of the chemical in the search bar and see if there is already a item with the same label. Snipre (talk) 14:03, 2 December 2017 (UTC)Reply[reply]

@Snipre: : Hi Snipre, thank you for your comment on my edit. I always check if a chemical already exists, but apparently I overlooked this. I am currently working as a PhD candidate with the group of Egon Willighagen and I am adding identifiers from other databases, which are used in pathways from WikiPathways, but do not link to Wikidata items yet (we want to use Wikidata as a unifying identifier). Especially for isomers, I sometimes find that identifiers from other databases are not correctly added to Wikidata. Following the suggestion from Sebotic above, I will add the "part of/has part" for the corresponding isomers. I would also like to suggest to do a query on overlapping identifiers from other databases (such as canonical/isomeric smiles) from time to time, to check if different Wikidata items have the same identifier listed. This would be a good step in chemical compounds curation, because I have seem some inconsistencies before. Do you know of any other curation efforts, specifically for chemical compounds?


Are you sure that sodium chloride (chemical compound) is a mixture (material made up of at least two compounds/elements)? Wostr (talk) 13:10, 3 January 2018 (UTC)Reply[reply]

@Wostr: Hi Wostr; When the salt is in the solid form, it is a material made up of two elements, right? In the fluid form, it is definitely a mixture (H2O & Na+ & Cl-)... Or do you suggest to model salt in another way (Perhaps we should add a suggestion at ). Denise Slenter (talk) 13:15, 3 January 2018 (UTC)Reply[reply]
I think that mixture is not the right term here. In the fluid form (melted) NaCl is still a compound (there is no water, only Na+ and Cl- ions). But NaCl can be mixed with water and it will form a mixture (then it is not sodium chloride (Q2314) but saline water (Q496805) or something similar). Wostr (talk) 13:29, 3 January 2018 (UTC)Reply[reply]
:@Wostr: So then we should label salts as ionic compounds? Maybe I'm confused with the Dutch terminology (where salts are called a mixture of elements...). Do you know how salts are being labeled now in Wikidata? Is there consensus on this? And then there is the example that you give, melted salt is different then a salt solution; very true indeed... I can imaging that these will have different identifiers as well... I brought up the issue of salts, since we(@Egon Willighagen: and me) found several identifiers for compounds in the biological pathways of Wikipathways, which were salts; however these salts will never be measured in a biological sample. It would be very helpful for us (and other researchers using Wikidata I hope) if we could check if the identifiers we have now, have been labeled as a salt in Wikidata.
I'll write about this at Wikidata:WikiProject_Chemistry/Proposal:Models, as there may be some language-specific differences, which can be hard to resolve here. Wostr (talk) 17:30, 3 January 2018 (UTC)Reply[reply]

Share your experience and feedback as a Wikimedian in this global survey[edit]

WMF Surveys, 18:57, 29 March 2018 (UTC)Reply[reply]

Reminder: Share your feedback in this Wikimedia survey[edit]

WMF Surveys, 01:39, 13 April 2018 (UTC)Reply[reply]

Subclass loop[edit]

Hi - your recent changes have made a loop - C16 DHLactosylceramide (incomplete stereochemistry) (Q37614694) is a subclass of beta-D-galactosyl-(1->4)-beta-D-glucosyl-(1<->1')-N-hexadecanoylsphinganine (complete stereochemistry) (Q27158043), which is a subclass of C16 DHLactosylceramide (incomplete stereochemistry) (Q37614694). Only one of those can be true, unless they are the same in which case they should be merged! ArthurPSmith (talk) 12:48, 24 August 2018 (UTC)Reply[reply]

Hi @ArthurPSmith:; thank you for checking so accurately ;) I've been doing a lot of manual curation last few days (for Wikidata IDs which have a tertiary identifier in common, but are not the same compound). I think I messed up on this one, will fix it asap. How did you find it? If there is a tool/script, I could use it next time after manual curation to check if I didn't loop anything :).DeSl (talk) 13:37, 24 August 2018 (UTC)Reply[reply]
The Ontology project has a few pages of "problems" like these loops that I check regular - see Wikidata:WikiProject Ontology/Problems. ArthurPSmith (talk) 20:22, 25 August 2018 (UTC)Reply[reply]

Ammonium cation[edit]

I think chemical compounds (neutral) should be separated from ions. We don't have established definition of chemical compound here on WD, but that's the way e.g. ChEBI uses and that's the definition (chemical compound = neutral) I often see in literature. Regards, Wostr (talk) 13:50, 26 August 2018 (UTC)Reply[reply]

Hi @Wostr:, thank you for checking my edits ;). Since the definition of chemical compound is also used for larger compounds with a charge (such as CDP-choline (1+) (Q421308)), and the definition reads: "pure chemical substance consisting of two or more different chemical elements", I added it for the ammonium cation ammonium cation (Q190901) . So I think the definition has to be revised, perhaps to "pure chemical substance consisting of two or more different chemical elements, which is not a cat- or anion" ?(since I think we don't see the larger compounds with a charge as a cat- or anion)... And if we change a definition, we should perhaps check which other items we combine with "instance of" and "subclass of", and revise the hierarchy between all these items? DeSl (talk) 07:00, 27 August 2018 (UTC)Reply[reply]
Hey :) I'm not checking your edits, it just popped up on my watchlist, as I have a few thousand items about compounds on it. There were some discussions about it, but... were inconclusive. The problem is we have a lot ions that have instance of = chemical compound, because of a large bot imports from various databases. Maybe there should be another item (a subclass of polyatomic ion (Q1064242)) that would cover every ion consisting of at least two different elements. Then it would be very easy to separate ions from chemical compounds or link them depending on discussion in the future (only by adding or deleting subclass of = chemical compound) – but I don't know how to even name such ion (complex ion is about something different; molecular ion = polyatomic ion, etc.). Wostr (talk) 08:45, 27 August 2018 (UTC)Reply[reply]

Amino acids[edit]

1. Labels should not start with capital letters (excepts some prefixes like L/D etc.). 2. Statements like L-glutamic acid (Q26995161)instance of (P31)DL-glutamic acid (Q181136) are correct (there is an agreement in WikiProject chemistry for such classification). Wostr (talk) 17:47, 28 March 2020 (UTC)Reply[reply]

@Wostr: Okay thank you for explaining. The reason I was making edits yesterday, was that there were inconsistencies in the proteinogenic amino acids. I only made changes to align the content. For example, if I want to sort on labels, the ones without a capital after the L- were not alphabetical. Furthermore, the links to "proteinogenic amino acid" and "L amino acids" where sometimes linked with "instance of", other times with "subclass of". Is there a specific section of the chemistry project I could look at, where the modeling rules for amino acids are explained? DeSl (talk) 09:23, 29 March 2020 (UTC)Reply[reply]
Right now we have only Wikidata:WikiProject Chemistry/Guidelines on which active users of this WikiProject agreed and we're trying to find items about 'group of compounds'/'group of stereoisomers' and adapt them to this model. About capital letters: in Wikidata labels shouldn't be capitalised with a few exceptions (names etc. + e.g. rare situations like locants and prefixes in chemistry). So it should be the other way around – change the ones with capital letters after prefixes. Wostr (talk) 16:54, 29 March 2020 (UTC)Reply[reply]