I am a tool developer with the Reactome team and I’ve recently added my tool to usegalaxy (not in iuc as we’d prefer to manage our tool in our own repo). I’d love to have some example datasets on usegalaxy for people to get started with, but can’t find the process to add a data library. Can someone point me in the right direction?
Hi @Paul_Wolfe
At the public UseGalaxy servers, the administrators control the Data Libraries. Since this is always a direct upload, contacting the server administrators directly is how it is usually done.
My immediate concern is that users might not know to look at a Data Library for example data, and if they find it, they then might not know how to use it! Other tool authors have hosted data at Zenodo, Github, some tutorial they hosted (in a dedicated “using our tool in Galaxy” section), or created a small GTN tutorial. Those are then all linked from the bottom of the tool form in the Help (where users look first for resources). This has the advantage of the data/instructions being accessible at every server that hosts the tool, and the exact data/instructions not being fixed/still under the control of the developer, since it the content is all just links.
But let’s see if others have ideas. This has come up a few times recently and I don’t think we had a set recommendation. Ping @wm75 or @bjoern.gruening – what do you think?
Then, for UseGalaxy.org specifically, I can help, but I wasn’t sure if this was for all UseGalaxy servers or not at this step? What did the others do?
Thanks @jennaj ! We also link to our github and user guide in the help section, and the reactome analysis tools site also has the examples. We were thinking this might just be a handy addition, especially if this is where users would “expect” data to be. But sounds like that may not be the case, so I’m curious to what the others think!
I’m curious too.
I did see the user guides and other resources. One idea would be to move up this
Visit the Reactome User Guide for detailed documentation about each tool.
to be right under the other first sentence in the Help. Then to add a sentence stating something like “Visit for example Galaxy usage and sample data!”.
Then on your guide, create one or two more sections for Galaxy. Deeptools is one example where that was done, see deepTools: tools for exploring deep sequencing data — deepTools 3.5.6 documentation. In here somewhere, maybe add something about checking the server’s Data Libraries as a way to alert people the data may be pre-loaded (this is what the GTN does).
Another idea is to promote your workflows to the iwc and host it here: https://iwc.galaxyproject.org/. Click on any of the workflows and review the lower right part of the view: “You can choose to run the workflow with sample data prefilled, or with your own data.” with a button to launch the example. Combined with the workflow descriptions and links to publications/resources – having this sort of automatic way to see what it all does is a very slick and easy way to get oriented, especially for non-technical researchers.
The iwc way is newer but might not reach as many users in the near term (but it will probably, eventually, be linked to tool forms using common tool_ids). But longer term – Galaxy will be “about the workflows” and ways to integrate the resources are a big focus, making this a forward-looking option.
Just some ideas. You could consider doing both! But let’s hear from the others, maybe I’ve missed another choice. Putting data in a library isn’t difficult. My concerns are that it will just “be there” and not really used since it is disconnected from the other resources and that it is a one-off task per server, distinct from the tool install. Would admins installing the tool have something in a README to add it in as a separate step? Those kinds of things. The option above makes the content available to anyone who installs it via Help, then at the UseGalaxy servers the IWC site layers in the full “demo” mode.