User:UreomiczBot/Human metabolites
Introduction[edit]
WikiPathways (Q7999828) is a database of biological databases. It contains many metabolites, a good number of which are not currently not found in Wikidata (Q2013). I have been previously adding metabolites using Bioclipse (Q1769726) and QuickStatements (Q20084080). However, that does not scale. Moreover, of metabolites that are present in Wikidata, not all of them have enough information to chemical identify them accurately.
The objective of this task is two-fold:
- annotate existing compounds in Wikidata as found in in a human (current)
- add metabolites not yet in Wikidata to Wikidata (implemented but currently failing with an API error about URLs in labels (which is not the case))
The first is done by adding a found in taxon (P703) statement with the value Homo sapiens (Q15978631). The provenance will express that the statement stated in (P248) WikiPathways (Q7999828), when it was retrieved (P813), the WikiPathways ID (P2410) it was stated in, and the reference URL (P854).
Current Scope[edit]
The set of entities maintained by this bot are determined based on their PubChem CID (P662) and the InChIKey (P235).
At present, the bot is limited to metabolites from pathways from the species Homo sapiens (Q15978631) with an PubChem CID (P662) (this is currently 151 metabolites).
Items maintained by this bot[edit]
- entities with a PubChem CID or InChIKey that are metabolites in WikiPathways
Metabolite properties currently maintained by this bot for these items[edit]
| Property | Datatype | Expected value
(if not listed, see property definition) |
|---|---|---|
| found in taxon (P703) | Item | Currently should only include Homo sapiens (Q15978631) |
| InChIKey (P235) | external-id | Added only for new compounds |
| InChI (P234) | external-id | Added only for new compounds |
| canonical SMILES (P233) | string | Added only for new compounds |
Metabolite properties PLANNED for this bot[edit]
| Property | Datatype | Expected value
(if not listed, see property definition) |
Note |
|---|---|---|---|
Data sources[edit]
The bot retrieves its content from the following sources (with a SPARQL query which can be found in the bot source code):
- WikiPathways (Q7999828) (disclosure: User:Egon Willighagen is an upstream curated)
Prototype items[edit]
- 4'-hydroxydiclofenac (Q26690136) (specially created to test the model and bot, together modeled with User:Andrawaag)
Bot test edits[edit]
- sn-glycerol 3-phosphate(2-) (Q26711901) (New item! I got it working in the end by removing all URLs, including the calendar model(!), so no 'reference URL' and 'retrieved' statements in the reference for now!)
- cortisol (Q190875) (diff) (diff of an bot update)
- Serine (Q7454846) (diff)
- lysophosphatidic acid (Q2823281) (diff)
- clarithromycin (Q118551) (diff)
An earlier version of the bot, using different data model made the following changes: 4'-hydroxydiclofenac (Q26690136) (diff), flutamide (Q418669) (diff), cortisol (Q190875) (diff),Serine (Q7454846) (diff), clarithromycin (Q118551) (diff), silibinin A (Q425702) (diff), and lysophosphatidic acid (Q2823281) (diff).
Bot approval[edit]
The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section. A summary of the conclusions reached follows.
UreomiczBot 1[edit]
UreomiczBot (talk • contribs • new items • new lexemes • SUL • Block log • User rights log • User rights • xtools)
Operator: Egon Willighagen (talk • contribs • logs)
Task/s: The task is to annotate chemical compounds in human metabolic pathways found on WikiPathways as such and to add chemical compounds in those pathways that are not yet in Wikidata. The latter is determined based on the PubChem CIDs given in WikiPathways and the matching InChIKey as provided by PubChem.
Code: The code is available from the feature/metabolites branch on BitBucket.
Function details: The code makes use of the User:ProteinBoxBot code base as found in the wikidatabots project of the Su lab also on BitBucket.
The code first aggregates information:
- it retrieves PubChem CIDs for chemical compounds marked "Metabolite" in human pathways in the WikiPathways database
- it looks up the SMILES, InChI, and InChIKey using the PubChem PUG REST API
- it checks Wikidata for an existing entry with either the PubChem CID or the InChIKey (in that order, consistency is not checked!)
- if not existing, it will create (current blocked by an API error) a new item, with label and PubChem CID (with source reference pointing to WikiPathways), and SMILES, InChI, and InChIKey (with source reference pointing to PubChem) and instance of (P31)-ing the entry as a 'chemical compound' 'found in taxon' 'Homo sapiens'
- if it does exist, it will only instance of (P31) the entry as a 'chemical compound' 'found in taxon' 'Homo sapiens'
The User:Ureomics/Human_metabolites task page has further information, including on the current test runs. Questions, suggestions, and other forms of feedback are most welcome.
--Egon Willighagen (talk) 13:30, 27 August 2016 (UTC)
- Normally a bot must have the word "bot" in the username: «Bot accounts are generally named after either their operator or their function, combined with the word "bot"» Can you change the name? --ValterVB (talk) 13:54, 27 August 2016 (UTC)
- @ValterVB: Yes, I know. I actually did that, but apparently forgot to write down the password and did not give it an email address (or so it seems). I asked support on #wikidata, but they could not help. So, in fact, User:UreomicsBot exists, but I currently don't have access to it. Once some admin restored my access, I can indeed rename the account. --Egon Willighagen (talk) 14:16, 27 August 2016 (UTC)
- If you have forgot the password none can help you, but you can create a new user (ex. UreomicsBot2 or Ureomics2Bot) --ValterVB (talk) 16:18, 27 August 2016 (UTC)
- OK, done: User:UreomiczBot. I will make a few test edits with that one too. --Egon Willighagen (talk) 18:01, 27 August 2016 (UTC)
- @ValterVB: What is the expected protocol now? Leave this bot (task) request with the original name, open a new request, or rename the bot name mentioned in this request? --Egon Willighagen (talk) 18:38, 28 August 2016 (UTC)
- @Egon Willighagen: You can move this page the to the new name. If you have problem ping me. --ValterVB (talk) 19:00, 28 August 2016 (UTC)
- Done! --Egon Willighagen (talk) 19:09, 28 August 2016 (UTC)
- @Egon Willighagen: You can move this page the to the new name. If you have problem ping me. --ValterVB (talk) 19:00, 28 August 2016 (UTC)
- If you have forgot the password none can help you, but you can create a new user (ex. UreomicsBot2 or Ureomics2Bot) --ValterVB (talk) 16:18, 27 August 2016 (UTC)
- @ValterVB: Yes, I know. I actually did that, but apparently forgot to write down the password and did not give it an email address (or so it seems). I asked support on #wikidata, but they could not help. So, in fact, User:UreomicsBot exists, but I currently don't have access to it. Once some admin restored my access, I can indeed rename the account. --Egon Willighagen (talk) 14:16, 27 August 2016 (UTC)
Support I really welcome this bot. It is using the same codebase as the ProteinBoxBot. --Andrawaag (talk) 18:09, 27 August 2016 (UTC)
Comment User:Wostr pointed out on my Discussion page that found in taxon (P703) can be used directly on the Wikidata entry being added/edited, so, instead of instance of (P31). The provenance model can remain the same then. If there are no objections, I will update the proposal accordingly tomorrow. --Egon Willighagen (talk) 18:49, 27 August 2016 (UTC)
Support I just went through today's test edits, which look good and useful to me. Agree that it's better to use found in taxon (P703) as a property rather than qualifier. --Daniel Mietchen (talk) 13:01, 29 August 2016 (UTC)
Implementation[edit]
The bot code is open source and available for inspection. It is implemented in Python, based on the User:ProteinBoxBot source code and is intended to be deployed as a cron job. The current operation is manual, however.