4
$\begingroup$

Hail from Stack Overflow!

Over on Meta Stack Overflow, the community has recently noticed and become grumpy about the questions in Stack Overflow's [cytoscape] tag. Many of them are purely about how to do things in the Cytoscape UI and seem to contain no programming-related problem (or else are programming-related merely in the tangential sense that they involve using Cytoscape to work with data files previously created by code the author wrote).

For better or worse, these questions are off-topic on Stack Overflow by policy, and the community there appears to me to be gearing up to close them en masse. Several of the newest questions have been closed in the last 24 hours, and as a high-rep user on Stack Overflow, I can see that more besides have accrued close votes and are likely to be closed too.

Most of these questions don't seem bad to me, at least at a glance; they're just not in-scope for Stack Overflow. The tag also has a dedicated answerer in the form of Cytoscape team member John "Scooter" Morris, and many of the questions have answers (often written by Scooter) that seem - at least upon a superficial inspection by me, a non-expert - to be useful.

Would such questions be in-scope on Bioinformatics Stack Exchange? I am not a bioinformatician, but as far as I can tell, Cytoscape is a tool originally created for, and almost exclusively used for, visualising biomolecular data by bioinformaticians. Are questions about using it therefore generally on-topic here (even if they are just software usage questions without either a programming element or a scientific element to them)? (I note that this site has three questions already, all of which are just about installation or usage, and none are closed, but I don't want to infer too confidently from that such questions are definitely welcome here.)

It seems to me like the bioinformatics community - and thus the Bioinformatics Stack Exchange community - probably has an interest in what happens to those questions. I'd like to tentatively suggest a few things that we could ask the mods on Stack Overflow to do - if this community's approves of them. (Obviously, a lot of the points below will automatically be a bad idea if you decide these questions are generally off-topic here, too.)

  • Migrate all [cytoscape] questions that are off-topic on Stack Overflow and new enough to migrate from Stack Overflow to Bioinformatics, and reopen them
  • Ask a CM if there's a way to get an exception to the 90 day site migration limit and migrate all older off-topic questions that the Bioinformatics community thinks belong here. (I realise you won't want to be a dumping ground for Stack Overflow's garbage and will want to get to filter these; someone from this site familiar with Cytoscape could skim all the Stack Overflow [cytoscape] questions and sort them into "migrate" and "leave on Stack Overflow" bins.
  • Agree a stock comment to leave on closed [cytoscape] questions noting that questions about Cytoscape that are not about programming should be posted on Bioinformatics Stack Exchange or Biostars instead of on Stack Overflow, and post that comment under all the [cytoscape] questions that get closed on Stack Overflow. (That way, at least interested users who come across old questions on Stack Overflow will have a signpost pointing them to a place where their further questions will be welcome.)
  • (If we can't migrate them or you don't want them migrated here en masse, then...) put a historical lock on off-topic Cytoscape questions (or at least on good ones, per some way of determining "good"), instead of just closing them, so that they won't be susceptible to downvoting and auto-deletion over time.

Regardless of any of the above, it seems likely to me that someone from the Stack Overflow community will in due course reach out to Scooter and the Cytoscape team and ask them to update their documentation on where to ask for help, which currently tells them to use Stack Overflow for "questions about software installation, operation and troubleshooting". When that happens, do you want us to mention Bioinformatics Stack Exchange? Would you like those Cytoscape docs to be directing users here? (I note that Biostars is already suggested there, which makes figuring out precisely what to suggest a little more complicated.)

Finally, we could also modify the tag excerpt and tag wiki on Stack Overflow to suggest Bioinformatics Stack Exchange (and Biostars?) as an alternative place to ask about Cytoscape where non-programming questions would be on-topic - again, if you'd like us to do so.

What say you? I've proposed lots of possible actions above that I suspect would give this situation a better outcome, but most of them are things that would need this community's approval and consent before anyone from Stack Overflow does them. Are they good ideas?

$\endgroup$
9
  • $\begingroup$ Thanks @MarkAmery. It sounds to me that the problem is with Cytoscape documentation, via unauthorised commercial use of SO platform. It is ettiquette to ask via meta, like you've done. Redirecting future posts is reasonable in my opinion, but let's discuss it. I think overriding 90 day retrospect migration is unlikely, thats up to SO, because the problem in the first instance was Cytoscape documentation. Do you have any interests to declare about Cytoscape? $\endgroup$
    – M__ Mod
    Commented Apr 27 at 16:42
  • $\begingroup$ "I think overriding 90 day retrospect migration is unlikely, thats up to SO" - well, in the first instance it's up to this community. There's no way anyone from the Stack Overflow mod team is even going to raise the possibility with the CMs unless you guys say you'd like it to happen. $\endgroup$
    – Mark Amery
    Commented Apr 27 at 17:15
  • 1
    $\begingroup$ "Do you have any interests to declare about Cytoscape?" - no, none. I'm just a Stack Overflow user who doesn't like valuable content getting purged from the internet or its creation prevented, nor skilled and altruistic people (like Scooter seems to be) getting their work rebuffed, and would like to see if there's a way to avoid those outcomes while still satisfying the rules of SO and the network (... and while not antagonising you guys on Bioinformatics, either!). $\endgroup$
    – Mark Amery
    Commented Apr 27 at 17:15
  • 1
    $\begingroup$ It does seem to be blatantly off topic on SO. I think you can certainly go ahead and edit the wiki suggesting this site for biological questions and either U&L or SU for the rest. I'd wait for some more reactions here before triggering any further action (although our Meta here isn't really very active). $\endgroup$
    – terdon Mod
    Commented Apr 27 at 17:22
  • $\begingroup$ @terdon Yeah, I noticed the fairly inactive Meta. I was loathe to suggest this until you also brought it up, but if you want to get some eyeballs here and think it's appropriate to do so, you could always feature this question for a couple of days so it shows up in the sidebar on the main site. (Doing so routinely for questions you deem worthy of community attention might be a good policy on a site with an inactive Meta.) $\endgroup$
    – Mark Amery
    Commented Apr 27 at 17:27
  • 1
    $\begingroup$ Yep, I was thinking the same. It's the only way to get some eyeballs on inactive metas indeed. Done now. $\endgroup$
    – terdon Mod
    Commented Apr 27 at 17:35
  • 4
    $\begingroup$ @M__ By the way, something I missed before but seems worth noting - you mention "unauthorised commercial use" of SO, but Cytoscape is open source and it's not obvious to me that the Cytoscape dev team or The Cytoscape Consortium (a nonprofit) are properly understood as "commercial" orgs at all, nor that Scooter - an academic - personally receives any compensation from them. I don't think money is a pertinent part of the story here, so let's be cautious about saying anything potentially untrue about these people - we're going to want to collaborate with them on friendly terms, after all. $\endgroup$
    – Mark Amery
    Commented Apr 27 at 17:40
  • $\begingroup$ Thanks @MarkAmery it all looks good. Open source project focused on biological molecular data and no conflicts of interest. $\endgroup$
    – M__ Mod
    Commented Apr 27 at 22:19
  • $\begingroup$ An update: I've emailed Scooter. Now we wait and see what happens next! $\endgroup$
    – Mark Amery
    Commented Apr 30 at 18:43

3 Answers 3

Reset to default
5
$\begingroup$

First off, thank you for taking the time to post here and ask. That is always appreciated!

My take would be that the questions will indeed be welcome and on topic here but only as long as they are about using Cytoscape to analyze biological data. In principle, the tool itself is just a general utility to analyze network data. Yes, it is very much used in bioinformatics, but there is no reason someone can't use Cytoscape to analyze any kind of network/graph data, and we don't want questions about counting Facebook connections or whatever here.

Installing it I think would be OK. My thinking is that since it is indeed a standard workhorse for anyone analyzing biological networks, and that is a large chunk of the community, many bioinformaticians might end up needing help installing it. More importantly, even if the OP isn't working in bioinformatics, any installation issues we can resolve are likely to help someone else who is working in the field, so I would argue such questions would be helpful here.

I would therefore propose migrating questions about installation and general use, but not those that explicitly focus on data that are not from the world of bioinformatics. For such questions, those that are about using Cytoscape on non-biological data, I would suggest posting either on Unix & Linux (assuming the OS the software is used on is a *nix flavor), or Superuser if the tools is being run on Windows.

In other words, if the question is about installing the tool or using it to analyze biological data, send it to us. If the question is about using it but not to analyze biological data, send it to U&L or SU.

I don't think SE would be willing to change the 60 day limit and that would likely require dev effort and manual DB interventions so I doubt it is worth it. I would focus on the newer questions instead.

Please note that this is my personal take as a user. Don't be distracted by my diamond, if others disagree and have better proposals, they may well be right. Let's see what the community has to say.

$\endgroup$
4
  • 1
    $\begingroup$ FWIW, any CM should be able to do a mass migration IF that was desired - the tool for this was originally created to allow CMs to move questions when closing a failed site, and just takes a list of question IDs. Also... Random trivia: in years past, I used cytoscape.js for visualizing networks of users on most of these sites when investigating voting fraud, and made good use of some questions on SO when implementing that; certainly wouldn't expect a blind mass migration to be appropriate. $\endgroup$
    – Shog9
    Commented May 2 at 16:50
  • $\begingroup$ Thanks @Shog9, that's good to know. I had thought it was more of an actual SQL command and so would require a dev since most of the current batch of CMs aren't programmers so are, very understandably, leery of running manual commands. I am more interested in whatever tool you have that brought this post to your attention though! I didn't even need to mention your name three times or anything :P $\endgroup$
    – terdon Mod
    Commented May 2 at 17:38
  • 1
    $\begingroup$ Was browsing MSO and saw mention of Cytoscape - still have fond memories of working with the JS lib, so followed the trail to see where it led. $\endgroup$
    – Shog9
    Commented May 2 at 17:49
  • $\begingroup$ @terdon Since it was me who originally raised the idea, I just want to explicitly note that I, personally, am not gonna follow up on getting questions migrated. I figure if it's gonna happen, it needs an SME to look over the questions and figure out which ones are off-topic-on-SO-but-on-topic-on-BSE & construct a list of Qs to migrate; then it ought to get run it past either the Stack Overflow mods via a mod chat or the wider community via Meta; and finally a mod needs to raise the request to a CM. As a non-mod and non-SME, I'm not up to any of that, so I wish you all well and bow out here. $\endgroup$
    – Mark Amery
    Commented May 2 at 21:37
2
$\begingroup$

Cytoscape looks like a well-developed open source tool, and we have had similar requests from a couple of developers in the past. I'm happy for questions to be migrated over, even if they don't explicitly mention a bioinformatics project, as long as they relate to a specific problem.

This question is an example of a generally-applicable Cytoscape question that isn't explicitly about bioinformatics (which I'd be happy to be migrated here):

https://stackoverflow.com/questions/48177816/how-to-change-nodes-color-individually-in-cytoscape/48326939#48326939

It is applicable to a general Cytoscape problem, and answers would help people who use Cytoscape for its intended purpose.

Updating the documentation for Cytoscape would also be a good idea.

FWIW, I don't as a general rule modify answers or posts to remove chatty text (despite this being against the official Stack Exchange policy). If the answers are otherwise good quality, I don't mind gratitude or other similar embellishments.

$\endgroup$
1
  • 1
    $\begingroup$ Agreed, if the question is broad enough that it could well be applied to a bioinformatics issue, we should welcome it. I would still close questions that are explicitly not about a biological question, say if the OP is trying to plot connections between movies and actors or whatever, but if they don't make it explicit, and the question can be asked when working in bioinformatics, bring it on. $\endgroup$
    – terdon Mod
    Commented Apr 28 at 15:30
0
$\begingroup$

Okay so in summary from the Mod side of things, the answer is 'yes' to all questions. It's a well-developed open-source project focused on bioinformatics with the OP having no conflict of interest. The singular request is to update the documentation for this site.

Lets wait to see what other members might want to say. Whether we've anyone knowledgeable about Cytoscape is a different question.

$\endgroup$
3
  • 1
    $\begingroup$ I expect we have many people knowledgeable about cytoscape. It is a standard workhorse for network analysis, so anyone who's worked on PPI networks or gene regulatory networks is very likely to have a good working knowledge of cytoscape. I've used it a lot myself, and I doubt I'm alone. $\endgroup$
    – terdon Mod
    Commented Apr 28 at 15:28
  • $\begingroup$ Okay thanks @terdon, I just do this de novo via graph theory. $\endgroup$
    – M__ Mod
    Commented Apr 28 at 18:33
  • 2
    $\begingroup$ Hi all -- Cytoscape trouble-maker here. Sounds good to us! We've updated our documentation to point to bioinformatics.stackexchange.com and I will now include that in my weekly triage. It would be really helpful if we could migrate as many of the Stack Overflow [cytoscape] but not [cytoscape.js] tagged posts over to here to provide some context. By the way, any questions about network biology, even if they don't explicitly involve Cytoscape are welcome. $\endgroup$
    – Scooter Morris
    Commented May 2 at 16:55

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .