User:ProteinBoxBot/Drug items

Introduction

The ProteinBoxBot maintains information about Genes, Diseases and Drugs in Wikidata. The entities in these three domains are maintained by different corresponding sub-processes of the main bot.

This objective of the Drug sub-bot is to add and maintain Wikidata items for all drugs relevant to human health.

Intended Scope

FDA-approved drugs and drug combinations.

Items maintained by this bot

The following Wikidata query will retrieve all items currently maintained by the bot. It retrieves all items that are both instances of a pharmaceutical drug (CLAIM[31:12140) and have a Drugbank ID (CLAIM[715]).

• Autolist (human) CLAIM[31:12140] AND CLAIM[715]
• API query (machine) CLAIM[31:12140] AND CLAIM[715]

Bot test edits

• Calcidiol
• Biotin
• Aspartame
• Aspartame
• Menadione
• Ampicillin
• Abciximab
• Cyclosporine
• Palivizumab
• Famciclovir

In order to query all bot test edit items, please use CLAIM[31:12140] AND CLAIM[715]

Prototype items

• Vemurafenib
• Warfarin
• Ipilimumab
• Atovaquone
• Dolutegravir
• Phenacetin
• Spironolactone
• Metformin
• Oseltamivir
• Rosuvastatin
• Clavulanic acid
• Naloxone
• Menotropins
• Pipobroman
• Mesalazine
• Guanadrel
• Urofollitropin
• Natalizumab
• Ampicillin
• Famciclovir
• L-Carnitine
• Mitotane

Data sources

The data sources for this effort are the open databases National Drug File, DrugBank, PubChem and ChEMBL. In order to determine the list of compounds currently approved by the FDA, Drugbank is used, it also provides a set of basic identifiers. In order to aquire more data, the RDF API auf PubChem is used to acquire the PubChem ID and MeSH ID. The National Drug File REST API is used to get the FDA UNII and ChEMBL is also queried directly for the ChEMBL ID.

Output

The bot will be able to add new drugs appearing in its source databases to Wikidata. It will also modify existing Wikidata drug items if new information becomes available in the mentioned data sources. The data added for a Wikidata drug item will initially comprise, additionally to labels and aliases, the following

Properties edited by the bot

Property	Description	Datatype
Property:P31	instance of	item
Property:P636	route of administration	item
Property:P267	ATC code	string
Property:P231	CAS registry number	string
Property:P486	MeSH ID	string
Property:P672	MeSH Code	string
Property:P662	PubChem ID	External ID
Property:P661	ChemSpider ID	External ID
Property:P652	UNII	External ID
Property:P665	KEGG ID	External ID
Property:P683	ChEBI ID	External ID
Property:P274	chemical formula	External ID
Property:P715	Drugbank ID	External ID
Property:P592	ChEMBL ID	External ID
Property:P233	SMILES	string
Property:P234	InChI	string
Property:P235	InChIKey	string
Property:P2275	Word Health Organisation International Nonproprietary Name	Monolingual text
Property:P657	RTECS Number	string
Property:P2115	NDF-RT ID	External ID

Implementation

The bot is split in two parts, the drug data aggregator and the actual drug bot which uses the aggregated data to write to Wikidata. The bot code is open source and available for inspection.

Bot approval

Bot approval discussion August 2015: Wikidata:Requests_for_permissions/Bot/ProteinBoxBot_4

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.

 Approved--Ymblanter (talk) 09:27, 12 August 2015 (UTC)

---

ProteinBoxBot 4

ProteinBoxBot (talk • contribs • new items • new lexemes • SUL • Block log • User rights log • User rights • xtools)
Operator: Sebastian Burgstaller (talk)

Task/s: ProteinBoxBot 4 will add new items and improve/maintain existing items on pharmaceutical drugs and drug like compounds and substances (small molecules and biologics).

Code: The code is divided into two parts. A core class which handles all Wikidata API related tasks [1] and a pharmaceutical drug data specific part [2].

Function details: The core class handles all Wikidata API related tasks and specifically takes care to select the correct item for writing data to. This module is used by all subtasks the ProteinBoxBot performs and is called by the resource specific part of the bot. The latter takes care of aggregating pharmaceutical drug data information from a set of well established scientific databases, e.g. DrugBank, PubChem, ChEMBL. A full list of properties maintained by the bot and the data sources used is available at the bot page [3]. The bot first aggregates the data from the listed resources and then writes the aggregated data to Wikidata. As a starting set, the bot will work on all FDA/EMA approved or withdrawn drugs listed in Drugbank. For each value, the data source is added as a reference and therefore very reliably maintains provenance. The bot will run at least weekly and will therefore maintain a very up-to-date list of pharmaceutical drugs and the corresponding data. --Sebotic (talk) 09:18, 28 July 2015 (UTC)

 Comment: 140 successful test edits have been performed. Sebotic (talk) 22:30, 11 August 2015 (UTC)

•  Support. Would be awesome to have the bot working on adding pharmaceutical drugs, next to the genes and diseases Andrawaag (talk) 15:21, 28 July 2015 (UTC)
•  Support. --I9606 (talk) 16:31, 28 July 2015 (UTC)
•  Support. I see a huge need op CCZero data to support pharmaceutical research of various kinds, and I really think getting more of this data into Wikidata will hugely benefit the field. Moving this data into Wikidata will create an open platform for linking related data, open ways for open community curation, create a bigger Linked Open Data network (e.g. link to PubChemRDF), etc. I really hope to see this move forward. Egon Willighagen (talk) 19:58, 28 July 2015 (UTC)
•  Support Andrew Su (talk) 16:04, 29 July 2015 (UTC)
•  Support --Daniel Mietchen (talk) 07:56, 12 August 2015 (UTC)
• @Ymblanter: Can this now be approved?--GZWDer (talk) 09:14, 12 August 2015 (UTC)

---

The above discussion is preserved as an archive. Please do not modify it. Subsequent comments should be made in a new section.