Template talk:Infobox drug/Archive 9

Integrate improvements and corrections

This edit request has been answered. Set the |answered= or |ans= parameter to no to reactivate your request.

Hello,

I have made improvements and corrections to this template at en.wikipedia.org/w/index.php?title=User:Nnemo/Drugbox&oldid=478751180. Please integrate this in the template. I was going to do it myself, but it's locked. Here are the changes.

Thanks,

--Nnemo (talk) 10:03, 25 February 2012 (UTC)

Hi Nnemo, and thanks for your work! I copied your version to

Edit protected

}} on its talk page, followed by a short explanation.

What is your reason for replacing simple spaces with {{sp}}? --ἀνυπόδητος (talk) 12:05, 25 February 2012 (UTC)

Otherwise, MediaWiki eats the spaces — which, by the way, is a bad thing.

--Nnemo (talk) 17:49, 25 February 2012 (UTC)

 Done

(Talk)

14:44, 26 February 2012 (UTC)

According to the DrugBank documentation, there are two types of DrugBank accession numbers, primary and secondary.

Currently there is a mix of primary and secondary accession numbers assigned to the DrugBank parameter in articles transcluding the Drugbox. Both primary and secondary accession numbers link to the same external DrugBank page, but the current usage is both inconsistent and confusing. The reason I bring this up is that I am trying to map Wikipedia drug articles to external databases (e.g., IUPHAR), and this is made more difficult by the inconsistent use of the DrugBank parameter. Hence I propose that only primary accession numbers are assigned to the DrugBank parameter.

Furthermore, is there any reason for secondary accession numbers to be displayed in addition to the primary number? The prefix of the secondary number does have the following significance:

The 4 letter prefix of the secondary accession number (APRD, EXPT, BIOD, NUTR) indicates the type of drug (APRD=approved small molecule drug, EXPT=experimental drug, BIOD=biotech drug, NUTR=nutraceutical or natural product).

If there is a strong desire to display the secondary in addition to the primary accession number, we could create a second parameter called DrugBank_Alt. For an example of how this would look, see the right hand side example in the

I don't see a reason for adding more numbers to the drugbox. The secondary numbers are accessible by simply clicking on the DrugBank link (which also should enable bots to access them if they are interested). Most people (like me until a minute ago) won't understand the abbreviations APRD etc., and this information should be pretty obvious anyway. (Small molecule, biotech product? Look at the structure or mass. Approved or experimental? Should be in the Legal status field.)

Unless any reasons for adding these numbers turn up, could perhaps a bot correct the secondary ones (hint, hint)? --ἀνυπόδητος (talk) 17:32, 8 April 2012 (UTC)

My preference is to keep things simple as possible and to only display the primary accession number. I just wanted to make sure I wasn't stepping on anyones toes before unleashing BogBot ;-) Boghog (talk) 17:41, 8 April 2012 (UTC)

Statistics if anyone is interested:

• Current number of Wikipedia articles that transclude the {{
  drugbox
  }}: 5084
• APRD prefix: 764 (15%)
• EXPT prefix: 21 (0.4%)
• BIOD prefix: 1 (0.02%)
• NUTR prefix: 0 (0%)

Boghog (talk) 19:46, 8 April 2012 (UTC)

 Done All secondary DrugBank accession numbers have now been replaced with their equivalent primary numbers. Boghog (talk) 19:15, 13 April 2012 (UTC)

Is there any way to change the solvent and temperature in the solubility data section? This is of use to the wider audience when concerning recreational drugs which users may wish to recrystallise.Testem (talk) 10:32, 14 April 2012 (UTC)

This template is hard wired to display the water solubility in units of 'mg/mL (20 °C)' if the solubility parameter is defined. This cannot be changed unless the template is modified. Furthermore according to

WP:NOTHOWTO. Those interested in recrystallizing drugs for recreational use must be a vanishingly small audience, so I doubt you will find a lot of support for modifying the solubility data section of this template. Boghog (talk

) 13:46, 14 April 2012 (UTC)

{{Drugbox/sandbox2
| drug_name = vemurafenib
| verifiedrevid = 432742697
| image = PLX4032_BRAF_inhibitor.png

<!--Clinical data-->
| tradename = Zelboraf
| licence_US = Vemurafenib
| pregnancy_US = D
| legal_US = Rx-only
| routes_of_administration = Oral
| Drugs.com = {{Drugs.com|pro|zelboraf}}
| MedlinePlus =

<!--Pharmacodynamic data-->
| Therapeutic_use  = [[Melanoma]]
| Biological_target = [[BRAF (gene)|BRAF]]
| MOA_text = [[Protein kinase inhibitor]]
| PDB_ligand = 032
| PDB_complex = 3og7

<!--Identifiers-->
| CAS_number = 1029872-54-5
| ATC_prefix = L01
| ATC_suffix = XE15
| PubChem = 42611257
| ChemSpiderID_Ref = {{chemspidercite|correct|chemspider}}
| ChemSpiderID = 24747352
| ChEMBL = 1229517
| UNII = 207SMY3FQT

<!--Chemical data-->
| C=23 | H=18 | Cl=1 | F=2 | N=3 | O=3 | S=1
| molecular_weight = 489.92 g/mol
| StdInChI_Ref = {{stdinchicite|correct|chemspider}}
| StdInChI = 1S/C23H18ClF2N3O3S/c1-2-9-33(31,32)29-19-8-7-18(25)20(21(19)26)22(30)17-12-28-23-16(17)10-14(11-27-23)13-3-5-15(24)6-4-13/h3-8,10-12,29H,2,9H2,1H3,(H,27,28)
| StdInChIKey_Ref = {{stdinchicite|correct|chemspider}}
| StdInChIKey = GPXBXXGIAQBQNI-UHFFFAOYSA-N
}}

IMHO, one major omission from the drugbox is information about the biological target and mechanism of action (i.e., "pharmacodynamic data"). This data could be added in a separate infobox as proposed above or incorporated directly into the drugbox (see example to the right). Is there support for adding this type of data directly to the drugbox? Boghog (talk) 21:01, 27 February 2012 (UTC)

• Note the "PDB drug" link provides a list of all the structures of proteins that have been complexed with this drug while the "PDB target" link contains information about how the drug interacts with its target (depicted as a LIGPLOT diagram).
• Also note this previous
  proposal
  to add class of drug or mechanism. The objection raised at that time was there was not room and it could not be explained properly. I agree that in some cases (especially for drugs with unknown mechanism of action), adding this data would not be appropriate, but in the case to the right where the drug has a well defined mechanism of action, I think the explanation is fairly clear.
• There is also this previous
  discussion where the argument against adding drug class information is that it duplicates information that should be contained in the lead and in addition, many drugs have more than one use (and by extension, possibly more than one mechanism of action). While this information should be contained in the lead (and also a mechanism of action section), often it isn't, Even if it is, it is often presented in an unstructured way this is difficult figure out what the mechanism of action is. If this information is in the drugbox, the reader can quickly figure out how this drug works. In cases where the MOA is not known or if the drug works through multiple mechanisms, I would agree it would not be appropriate to display this section in the drug box. Boghog (talk
  ) 06:33, 28 February 2012 (UTC)

Support adding these data to the drugbox. That's clearer and quicker to find (and edit) than if we had two boxes. We could mention in the drugbox documentation that more than 2 or 3 uses/mechanisms are discouraged. Perhaps we could link from the drugbox to the appropriate section in such cases (especially in long articles).

I think the header Pharmacodynamic data is suboptimal. This section contains data that are interesting for the general readers, and not many people know the word pharmacodynamic. What about Use and mechanism? Any better ideas? --ἀνυπόδητος (talk) 08:33, 28 February 2012 (UTC)

Briefly as I don't have a huge amount of time to comment right now. I think we are in danger of second guessing (incorrectly) what the user might want. By putting PDB info in a 'mechanism of action' section assumes that all drug structures in the PDB relate to the pharmacological target, and that users looking for protein-drug complexes require this information. The PDB structure might be of the target, or a resistance mechanism, or the last step of the biosynthesis pathway. Equally, this might be the information the user is looking for. Perhaps therefore the box name isn't quite right. For instance there are 16 structures containing chloramphenicol, most are not the ribosome.

If there are several structures of a drug in complex with the target, which ligplot is linked to?

Does the link to 032 page provide much more information than is available in the drug box anyway? — Preceding unsigned comment added by A2-33 (talk • contribs) 12:00, 28 February 2012 (UTC)

It is not clear to me how cases with more than one mechanism will be handled in this box. There may be structures in PDB for which the exact mechanism is not well understood. Also from discussion above it would be good to be clear n what exactly we want to achieve by adding the extra information.

• Do we want to give easy access to "mechanism" information?
• Do we want to give information about "interaction" of drug with macromolecules?

I also think that adding a link to PDB which only describes the molecule is not adding any extra information instead if we add a link to PDB structures e.g. http://www.ebi.ac.uk/pdbe-srv/PDBeXplore/ligand/?ligand=AIN for aspirin so then it shows all structures that have the molecule and the extra information about domains, assemblies etc. and something like http://www.ebi.ac.uk/thornton-srv/databases/cgi-bin/drugport/GetPage.pl?template=drug.html&drug_id=DB01614. This would be more useful in drugbox. A2-25 (talk) 13:00, 28 February 2012 (UTC)

Thanks everyone for your comments. Obviously the proposal needs to be fined tuned before it is put into production. I agree that "pharmacodynamics data" section heading is not ideal, but one advantage of this heading is that it is consistent with the "pharmacokinetic data" heading that is already contained in the drugbox (unfortunately this cannot be seen in the example to the right because this data is apparently not yet available for this drug). All the structures in the PDB containing a given drug may not relevant to the mechanism (but may be relevant to pharmacokinetic, e.g., drug metabolizing enzymes, transport proteins, etc.). The idea of the LigPlot link is that would be specific to the known target of the drug. The idea of the PDB ligand ID link is that it would provide links to all the structures containing the drug , some of which may be relevant to mechanism, others of which might be relevant to pharmacokinetics. In the current link, one has to click through to see the structures so that I agree that it is not ideal. I don't have much more time to comment at the moment, but I will resume experimenting in a day or two to try to incorporate your suggestions and address your concerns. Boghog (talk) 17:59, 28 February 2012 (UTC)

Thanks Boghog. A2-25 (talk) 21:48, 1 March 2012 (UTC)

• Per the above discussion, I have changed the title of the "pharmacodynamic data" section "mechanism & use" and moved the PDB links to the identifiers section and link changed the links so that list of all structures containing a particular PDB ligand ID are displayed. My original thought for including the PDB target links (as a LigPlot diagram which in turn linked to the PDB structure) in the mechanism section is that it provided better context for the links. Now the query link will provide a list of all structures complexed with the ligand, and it is up to the reader to determine whether the associated biopolymer is a target, metabolizing enzyme, transport protein, or has some other function. The changes can be seen in the infobox to the right. Does this look OK?
• Concerning mechanism/target data, I was (and still am) hesitant to include this type of data for drugs that have more than one mechanism of action because it would be difficult to locate and explain this data. However I recently noticed that the DrugBank database maintains a comprehensive list of targets for many drugs (see for example amiodarone targets). Furthermore one can download this data. Hence it would be possible for a bot to systematically add this data to drugboxes. In addition, this data could be displayed as a optionally collapsible table of Actions/Target pairs. Thoughts? Boghog (talk) 11:23, 3 March 2012 (UTC)

From my point of view (working in a pharmacy), "Mechanism:

antiarrhythmic (hERG)" or something (for amiodarone). Just ideas. --ἀνυπόδητος (talk

) 15:56, 3 March 2012 (UTC)

The new changes look good. I was looking at drugbank data and it would be good idea to add these links using a bot. A2-25 (talk) 17:27, 3 March 2012 (UTC)

Any progress on this front? A2-25 (talk) 20:55, 13 March 2012 (UTC)

OK. I will put on hold for the moment the "Mechanism & use" section but request that PDB ligand ID be added to the production version of the template (I cannot do this myself since I am not an administrator). As soon as this is done, I will fire BogBot to add and populate the PDB_ligand parameter. Boghog (talk) 21:17, 13 March 2012 (UTC)

Any progress on this front? Is this waiting for any data related issues or for an administrator to put it in production? A2-25 (talk) —Preceding undated comment added 20:29, 19 May 2012 (UTC).

The URL format for Daily Med has changed and links in Drugbox templates for DailyMedID entries no longer work.

Whywhenwhohow (talk) 22:43, 16 May 2012 (UTC)

Do you have an example where the link doesn't work? The DailyMedID link for furosemide seems to work. Boghog (talk) 22:52, 16 May 2012 (UTC)

The DailyMedID link fails in the template on these pages

Xyrem

The DailyMedID link fails on these pages, but not in the template

Lactated Ringer's solution, Quinine

,

Whywhenwhohow (talk) 18:11, 17 May 2012 (UTC)

Thanks for the examples. However even after reading the DailyMed documentation, it is not at all clear how to fix these links. It only the base url has changed, then this can easily be fixed in the template. If the accession numbers/format has also changed, this will be much more difficult to fix. Any ideas? Boghog (talk) 20:23, 17 May 2012 (UTC)

After doing some more digging, DailyMed provide some Web Services that might be useful in bot assisted fixing of the links. For example:

• http://dailymed.nlm.nih.gov/dailymed/services/v1/drugname/bimatoprost/spls.xml

returns an xml containing the setid which in turn can be used to create a link to the DailyMed product insert:

• http://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=6094c742-7cea-429d-8d21-b7cbf8bfe1bd

Unfortunately the drug name services doesn't seem to work for all INNs (for example http://dailymed.nlm.nih.gov/dailymed/services/v1/drugname/atorvastatin/spls.xml). Boghog (talk) 06:12, 18 May 2012 (UTC)

It appears that they are moving from using drugInfo.cfm?id= in the URI with a code parameter to using lookup.cfm?setid= in the URI with a GUID parameter and state that it is a permanent URL. If you use the drugInfo Minocin URL to view the page it displays a permanent link on the page using the lookup Minocin URL. Whywhenwhohow (talk) 12:03, 18 May 2012 (UTC)

The drugname search is more specific. You need to use atorvastatin calcium (or lipitor) instead of atorvastatin -

http://dailymed.nlm.nih.gov/dailymed/services/v1/drugname/atorvastatin%20calcium/spls.xml or

http://dailymed.nlm.nih.gov/dailymed/services/v1/drugname/atorvastatin+calcium/spls.xml or

http://dailymed.nlm.nih.gov/dailymed/services/v1/drugname/Lipitor/spls.xml

Use Penicillin V Potassium instead of Phenoxymethylpenicillin - http://dailymed.nlm.nih.gov/dailymed/services/v1/drugname/Penicillin%20V%20Potassium/spls.xml Whywhenwhohow (talk) 12:03, 18 May 2012 (UTC)

Unless DailyMed provides a more flexible searching mechanism, it will be very difficult to script a bot to find the appropriate links. DailyMed web services searching should be made more flexible. Searching using the INN should be sufficient. As it stands, these links will have to be manually updated. Boghog (talk) 12:54, 18 May 2012 (UTC)

A script could obtain the generic name from the drugs.com monograph and use that in the DailyMed web service search. How will a script decide which setid to use from the web service query results? For example, take a look at the results of the http://dailymed.nlm.nih.gov/dailymed/services/v1/drugname/Atorvastatin%20Calcium/spls.xml query. I suppose it could use the one with the latest published_date.

Please add a new parameter to the template, something like DailyMedSetID, that uses the permanent URL format so that Daily Med links can be added/updated in articles. Over time the existing IDs can be migrated to the new ones. Whywhenwhohow (talk) 22:45, 19 May 2012 (UTC)

Wikipedia drug articles are named after the INN. In principle a much more straightforward solution is to search for the INN, for example using this query:

http://dailymed.nlm.nih.gov/dailymed/advSearch.cfm?startswith=NAME:(*atorvastatin*)

and then extract the setids from the raw html. But then one needs to decide on which setid to use. I would argue the earliest rather than latest setid is most appropriate. This links to the original brand and formulation that is most likely to be approved for the widest indications. Furthermore product inserts are updated with new information, so even though the original version of the product insert may be out of date, it should be replaced with updates as they become available.

Alternatively why not just link to the INN search query results? If there are multiple brands, formulations, etc for a given drug, the decision on which product insert to link to is rather arbitrary and ideally should be left up to the reader. The disadvantage of course is the reader will have to click twice to get to a product insert. Boghog (talk) 21:41, 22 May 2012 (UTC)

This edit request has been answered. Set the |answered= or |ans= parameter to no to reactivate your request.

Please update from

There are some other changes in the sandbox by User:Boghog from earlier in the year. Can you confirm if these are ready for deployment as well? — Martin (MSGJ · talk) 16:17, 14 June 2012 (UTC)

As explained in more detail

here, the DrugBank_Alt parameter is not needed and I have removed it from the sandbox. The PDB_ligand is useful and I would appreciate if that parameter were included in the production version of the drugbox. Thanks. Boghog (talk

) 17:54, 14 June 2012 (UTC)

 deployed Thanks for clarifying — Martin (MSGJ · talk) 21:26, 14 June 2012 (UTC)

Please add "adverse effects" to the Drugbox. Under "Metabolism" would be suitable.--Taranet (talk) 08:25, 15 June 2012 (UTC)

The template documentation includes ...

| name =

...as one of the parameters for the template. But as far as I can tell, it doesn't do anything. However, the parameter...

| drug_name = 

...is used as the drugbox title. Is this just an error in the documentation? -- Ed (Edgar181) 18:25, 22 May 2012 (UTC)

Per this

discussion, the "name" parameter adds a {{Navbar}} to the top righthand side of the drug infobox. The name of the "name" is not very clear and probably should be changed to something like "path". In any case, this has not been documented. I will fix this. Boghog (talk

) 22:14, 22 May 2012 (UTC)

The "name" parameter was documented but was misplaced in the "clinical data" section of the documentation. The documentation for this parameter has now been moved

here where it should be easier to spot. Sorry for the confusion. Boghog (talk

) 22:38, 22 May 2012 (UTC)

OK, I understand now. Thanks for the clarification. -- Ed (Edgar181) 13:11, 23 May 2012 (UTC)

I've boldly modified the documentation to discurage users from using the name parameter. I have been removing it from at least a dozen drugboxes in the past few months to get rid of the "v t e" redlinks. It is only used correctly on two pages (Lisinopril, Atorvastatin), and Template:Drugbox/Metformin is not linked from Metformin since the latter has a chembox.

The links at the drugbox tops should either be hidden if no subtemplate exists, or the param should be removed altogether. Thoughts? --ἀνυπόδητος (talk) 13:37, 12 September 2012 (UTC)

The

GNF_Protein_box}} templates some time ago. I still think this is a good idea. A bot could systematically go through all the drug articles and create transcluded templates. If there is support for this proposal, I probably could convince BogBot to do this in the next month or two ;-) Thoughts? Boghog (talk

) 19:21, 12 September 2012 (UTC)

Support. Could your bot then periodically look out for new drugboxes to move to template space, to keep things consistent? --ἀνυπόδητος (talk) 06:42, 13 September 2012 (UTC)