MEpedia:Discussion forum and Help desk: Difference between revisions

From MEpedia, a crowd-sourced encyclopedia of ME and CFS science and history
m (Protected "MEpedia:Help desk" ([Edit=Allow only autoconfirmed users] (indefinite) [Move=Allow only autoconfirmed users] (indefinite)))
(→‎Questions: auto-signing section headers)
Line 62: Line 62:
:Citations from pubmed are fully working now - I haven't made any change since I last posted here. [[User:Canele]] [[User:JenB]] [[User:Pyrrhus]] if you find any examples of them not working please tag me and I will take a look. Canele's original [https://www.ncbi.nlm.nih.gov/pubmed/10990237 pmc ref] works for me on my user page. [[User:Notjusttired|notjusttired]] ([[User talk:Notjusttired|talk]]) 16:13, July 14, 2019 (EDT)
:Citations from pubmed are fully working now - I haven't made any change since I last posted here. [[User:Canele]] [[User:JenB]] [[User:Pyrrhus]] if you find any examples of them not working please tag me and I will take a look. Canele's original [https://www.ncbi.nlm.nih.gov/pubmed/10990237 pmc ref] works for me on my user page. [[User:Notjusttired|notjusttired]] ([[User talk:Notjusttired|talk]]) 16:13, July 14, 2019 (EDT)
{{done}} Now working.
{{done}} Now working.
===Auto-signing section headers===
Hi, starting a new section on a talk page automatically adds a signature to the header, with no possibility to remove it during the same edit. This is a very annoying feature. Can we get rid of it? [[User:Guido den Broeder|Guido den Broeder]] ([[User talk:Guido den Broeder|talk]]) 19:18, September 15, 2019 (EDT)

Revision as of 23:18, September 15, 2019

This MEpedia help desk is a centralized place to seek and give advice on the MEpedia project.

To ask a question or make a comment:

  1. Click on the "Add topic" tab at the top of the window, which will bring up the Source Editor. (If you are not familiar with using the Source Editor, see MEpedia:Discussion.)
  2. Then add your question or comment. Be sure to sign your question/comment with four tildes: ~~~~
  3. Keep an eye on your watchlist to see when someone has responded. (If you are not familiar with watchlists, see MEpedia:Watchlist.)

You can also raise questions on the discussion (talk) page of a specific entry or on another user's talk page. Since editor availability can vary, though, you may wish to ask your question here in order to get advice from whoever might see it soonest.

Note: Editors willing to offer guidance should please add this page to their watchlist. When a question/comment is resolved, mark it as ‘done’ with the “done” template tag, and collapse the discussion with the “collapse top” and “collapse bottom” template tags.

Questions[edit | edit source]

Request for science help re: gancyclovir[edit | edit source]

content hidden

I'm wondering if any more scientifically knowledgeable folks can help me with Gancyclovir. It's one of many pages that have no references at all and probably I'm best equipped just to tag them for cleanup, but since the entry only contains two claims, I thought I could track down refs to verify them. Ha ha. I found these studies which (if I understand them correctly) make opposing claims about its effect on microglial: "The antiviral drug ganciclovir does not inhibit microglial proliferation and activation" (2015) versus "Antiviral drug ganciclovir is a potent inhibitor of microglial proliferation and neuroinflammation" (2014). Is anyone available to help either revise the page (if I'm correct that maybe these are conflicting results that require qualification on the page) or confirm for me that the latter ref (or some other ref?) suffices to verify the entry's current claim that gancyclovir "also reduces microglial activation." Thanks much. Canele (talk) 17:45, 15 April 2019 (EDT)

Excellent question. I answer your question here. (Spoiler: it doesn’t reduce microglial activation after all.)
Pyrrhus (talk) 18:17, 15 April 2019 (EDT)
Oh, thank you so much Pyrrhus! Ok, I've updated both the gancyclovir and valgancyclovir. If you have a chance to look over and make sure I've presented it accurately, I'd be grateful! (Whew, glad we're fixing this! Def the kind of thing that argues for working hard to recruit and retain a lot more folks to comb the 2500+ pages for such issues...) Canele (talk) 20:57, 15 April 2019 (EDT)
Ah belatedly reading down the rest of that thread to see the follow-up STING paper from the first authors. I'm pretty well in over my head scientifically at this point. If you have guidance about if/how it should be incorporated, that'd be greatly appreciated but I don't want to assign you work (!); I can also just add it as a "Learn more" and eventually someone else can incorporate as far as explaining its implications. Canele (talk) 21:42, 15 April 2019 (EDT)
You covered the two papers very well in the two pages- succinctly, accurately, and objectively. I'll summarize the STING paper and then you can decide whether to include my summary in the page(s).
"The authors of the original 2014 paper published a follow-up paper which described in-vitro and in-vivo experiments where they observed evidence that ganciclovir, in sufficient doses, might stimulate a type-I interferon response in microglia. The authors attributed this response to a DNA-sensing protein called STING. Although this pro-inflammatory finding might appear to contradict the authors 2014 paper, the authors explain the contradiction by saying:

We show that GCV can exhibit dual function in microglia (Figure S10): in naïve state, GCV induces microglia to be “primed”; on the other hand, GCV reduces inflammation in active microglia.

"
Pyrrhus (talk) 00:07, 16 April 2019 (EDT)
Oh fab. For those of us with limited neuroscience background, can I take a stab at restating your explanation and you tell me if I understood it correctly?
The authors of the 2014 paper later published a follow-up paper (2017) based on in-vitro and in-vivo experiments using ganciclovir. This time they found a potentially pro-inflammatory response, specifically that ganciclovir, in sufficient doses, might stimulate a type-I interferon response in microglia. The authors attributed this response to a DNA-sensing protein called STING. While this result might appear to contradict their earlier anti-neuroinflammatory finding, they suggest it could indicate ganciclovir (GCV) "can exhibit dual function in microglia [...]: in naïve state, GCV induces microglia to be 'primed'; on the other hand, GCV reduces inflammation in active microglia."
If my lay-ified paraphrase gets it wrong tho, I say let's use yours; it does seem like a finding worth flagging for researchers even if some less scientifically literate readers (ha, yours truly) may not get it (although also you're definitely better qualified than I to judge whether the study rates mention. So whatever you think is best.) Canele (talk) 20:01, 16 April 2019 (EDT)
I‘ve made a few small changes to your text. Mainly, we have to be careful about our phrasing because there is currently a vicious fight over the very definition of “inflammation”. The old-timers stick stubbornly to the ancient greco-roman definition of inflammation, which does not recognize either the interferon response or neuroinflammation to have anything to do with what they consider to be “real” inflammation. Additionally, this is a hard paper to summarize because the authors themselves seem confused about their findings. Since you ask for my opinion, I don’t think their paper introduces anything new since their findings are perfectly consistent with the known toxicity of ganciclovir at high doses. But, in the interest of completeness, let’s include it. Hope this helps.
Pyrrhus (talk) 21:15, 16 April 2019 (EDT)
Ah thank you so much for your edits and your explanation Pyrrhus. This sounds like a good solution to me: the lay reader will def get the key take away that no, there's not firm evidence for this effect as the initial study suggested; and any scientist readers who want to get in the weeds will learn this study exists and they can go delve into its particulars if they wish. Cool! Thanks! Successful first use of the editor help desk! Canele (talk) 22:07, 16 April 2019 (EDT)

✓ Done Resolved and copied to relevant entry's talk page for future reference. Canele (talk) 20:24, 17 April 2019 (EDT)

Broken citation template?[edit | edit source]

content hidden

Hi all. I'm suddenly having trouble using the "Cite web" template. In particular I'm trying to add https://www.rheumatology.org/I-Am-A/Patient-Caregiver/Diseases-Conditions/Polymyalgia-Rheumatica as the reference for all the material I've added so far to the Polymyalgia rheumatica page I just started. But both in Visual Editor and in source code editing, I can't get the "Cite web" template to populate as a reference--it doesn't work at all in VE now and when I try source code, it adds all the info to the prose of the entry, rather than creating a footnote. Anybody have an idea what might be going on? I've used this extensively before (just for one example, see the HIV/AIDS page) and those references remain properly formatted so I am really stumped. Thanks for any advice! Canele (talk) 15:21, July 5, 2019 (EDT)

With Njt's help, I copied pre-formatted source code and successfully inserted the ref, but in VE, even the Manual tab is still not recognizing the "Cite web" template (even though Template:Cite web is there!), and it doesn't seem to vary based either on the entry I'm editing or on the link I'm trying to add (e.g. no luck adding https://www.britannica.com/science/glucocorticoid to the Glucocorticoid page even tho Britannica is whitelisted). Does anyone have any idea what's up with that? Canele (talk) 22:05, July 5, 2019 (EDT)
You’re not the only one to notice citation problems recently. It’s currently unclear to what extent the problems are on our side or on the side of the Citoid server. If it’s on our side , there are some things we can try. If it’s on the side of the Citoid server, then all wikis using that Citoid server should be seeing the same problems.
Pyrrhus (talk) 00:59, July 6, 2019 (EDT)
User:JenB has also reported pubmed links not being looked up, even ones that previously worked, and this appears to be related (they use the journal template). User:Kmdenmark User:Canele User:Pyrrhus What's happening is that Citoid, which generates the references, and/or the VisualEditor can't find the templateData so doesn't know what fields are in the template to fill out. This seems to have happened when the Citation, Cite web and other templates were accidentally then restored. I've copied the templateData from Template:Cite web/doc to Template:Cite web so the web one is working, but this isn't much of a solution since it means there's duplicate data there. Cite news still isn't working (Insert, Template, Cite news brings up no boxes to fill out). Looking elsewhere, a purge of the Citation json has been suggested (which basically means delete old versions) but I don't have access to do this. User:Kmdenmark You might see a purge option on MediaWiki:Citoid-template-type-map.json to click - if not try clicking once on this link to purge the page - I get "file not found" when I try, probably because the page is protected or I don't have edit access - it might need unprotecting first. (Purge instructions suggested at bottom of this page). I don't know how or why but the citoid tool or visual editor or both don't seem to see the templateData and/or some other values that exist. notjusttired (talk) 08:20, July 6, 2019 (EDT)
”Purging” refers simply to refreshing the cache, not removing old versions. You don’t need any special permissions to purge the cache. Since the MediaWiki:Citoid-template-type-map.json page hasn’t recently been changed, I don’t think that purging the cache will have any effect.
Pyrrhus (talk) 13:54, July 6, 2019 (EDT)
I think it's worth trying, although editing and saving without changing first is probably a good idea. Purging involves clearing old versions from the cache (temporary memory) in this case - but can also mean deleting old copies (I don't think mediawiki allows that). Normally to clear the cache online you just refresh the page, but I think it's the server cache that's the issue not our local cache. We are on web hosting do no physical server to restart. At the moment I'm trying to guess what might help since I don't know exactly where the problem is. notjusttired (talk) 15:23, July 6, 2019 (EDT)
Thank you both for the updates, it's very helpful to know at least it's not just me! Most of this punches well above my technical-skills weightclass so I'm afraid I don't have much to add, but if I'm reading this correctly, it looks like Citoid is enabled on en-Wikipedia, and I can confirm the ref issue is not happening there. Just in case that helps! Thanks again. Canele (talk) 16:45, July 6, 2019 (EDT)
Yeah, the “edit and save without changing” trick works for TemplateData, and may work on Template pages in general, but it might be less applicable in other scenarios. MediaWiki uses a number of different caches, but most of the time, people refer to the MediaWiki server’s Page Cache when they say “purging the cache”. To purge the server’s Page Cache for a specific page, simply append ?action=purge to the end of the page’s URL. (See https://www.mediawiki.org/wiki/Manual:Purge ). Hope this helps!
Pyrrhus (talk) 19:08, July 6, 2019 (EDT)
You probably already know this, but there is a known bug where TemplateData in the doc subpage doesn’t show up in the main Template page. I encountered this recently. Although there are workarounds, this problem seems to show up surprisingly often. I think it might be better practice for us to consider storing all TemplateData in the main Template pages, and remove all TemplateData from doc subpages. This way we won’t have to worry about this bug anymore...
Pyrrhus (talk) 21:20, July 6, 2019 (EDT)
Okay, the TemplateData for all CS1 templates ({{Cite x}}) and the CS2 template ({{Citation}}) has been moved from the /doc subpages to the main Template pages. This should clear up everything on our end. If problems continue, it's probably on the side of the Citoid server. (Unless our TemplateData is wrong!)
Pyrrhus (talk) 22:49, July 6, 2019 (EDT)
Thank you!! So after purging the cache on Glucocorticoid (just for instance), the Cite Web template does how populate in the Manual tab--however, automatic ref generating still isn't working even for sources we use all the time, like PubMed (I was trying to add https://www.ncbi.nlm.nih.gov/pubmed/10990237 without having to fill in every author manually). Just an update on where it stands, I know all this debugging is a ton of work. It's very greatly appreciated! Definitely affects how much content I am able to add, whether or not the refs can be largely automated or if all of it needs to be copied by hand. Thank you! Canele (talk) 23:14, July 6, 2019 (EDT)
User:Notjusttired The news and web citations are now being generated, plus doi links and everything but pubmed works as far as I know. I tried the purge suggested by User:Pyrrhus on all relevant modules and templates plus a trivial edit to the Cite journal template and a purge. I didn't know about the templateData bug but the solution sounds better than having duplicate templatedata like before. But pubmed generation still isn't working. I don't know how Citoid links with Cite journal, or Zotero - it's Zotero that generates the references apparently. Can't think what else to try. User:Canele Either a non-pubmed link to the article, or a doi link will generate a reference as a workaround. There's no question that if is just here, I'm also sure it was caused by the accidental delete of templates that were restored very soon after. notjusttired (talk) 17:17, July 7, 2019 (EDT)
So it sounds like we fixed the problem that we had on our side. Now there's just the remaining problem on the side of the Citoid server. The Zotero software that runs on the Citoid server must continually be updated to keep pace with the website changes on journal websites, pubmed website, etc. Sometimes the Zotero software on the Citoid server might fall behind, and will fail to automatically produce a citation based on a URL. In this case, you just have to try different URL's until you find one that can successfully be converted into a citation. Hope this helps.
Pyrrhus (talk) 20:56, July 7, 2019 (EDT)
Yes thank you both, that helps a lot and I agree that's where we are--just tried en-Wikipedia and encountered the same PubMed-specific problem. In this case it's an old text so I don't have a great option for different links to it (and predictably PMID is also down when PubMed URL is down) but that should be a good workaround for most things until they update Citoid/Zotero. A million thanks! Canele (talk) 21:04, July 7, 2019 (EDT)

✓ Done Resolved as best as we could. Pyrrhus (talk) 20:27, July 12, 2019 (EDT)

Citations from pubmed are fully working now - I haven't made any change since I last posted here. User:Canele User:JenB User:Pyrrhus if you find any examples of them not working please tag me and I will take a look. Canele's original pmc ref works for me on my user page. notjusttired (talk) 16:13, July 14, 2019 (EDT)

✓ Done Now working.

Auto-signing section headers[edit | edit source]

Hi, starting a new section on a talk page automatically adds a signature to the header, with no possibility to remove it during the same edit. This is a very annoying feature. Can we get rid of it? Guido den Broeder (talk) 19:18, September 15, 2019 (EDT)