galaxyproject / training-material

A collection of Galaxy-related training material
https://training.galaxyproject.org
MIT License
303 stars 891 forks source link

Mapping tutorial: loading IGV browser does not work #2349

Open jdavcs opened 3 years ago

jdavcs commented 3 years ago

Under "Visualization using a Genome Browser (IGV)", step 2 does not work, at least on usegalaxy.org: https://github.com/galaxyproject/training-material/blame/master/topics/sequence-analysis/tutorials/mapping/igv.md#L10-L11

If this is a know problem, maybe remove that step altogether? Or, at least, mark step 1 as recommended?

EDIT: this is also (or even more) relevant to this: https://training.galaxyproject.org/training-material/topics/introduction/tutorials/igv-introduction/tutorial.html

jdavcs commented 3 years ago

Also, maybe we need to add the following details on its local usage provided by @jennaj (all following text is the quote, it needs to be edited for the tutorial):

The trick with IGV local is that it not only has to be installed, but the app has to be open before you try to click into it.

They have indexes for many genomes that are easy to load. For CG in Galaxy, you have to make a custom build in Galaxy, give it a dbkey, assign that to your data in Galaxy, add the CG to IGV, give it the exact same dbkey, then can load multiple datasets. Otherwise, each dataset is a 1-off. Outside scope of that tuto but just in case someone asks, that’s how to do it. I have some Ghelp posts about it — and will promote that to an FAQ and maybe a GTN snippet, but not for this (too complicated, not needed)

Also, possibly, more relevant details here: https://help.galaxyproject.org/t/opening-unicycler-assemblies-with-igv-local/2653/4?u=jennaj

hexylena commented 3 years ago

We can update the tuto to use jbrowse? Then it's not an issue?

And EDIT: training.galaxyproject.org/training-material/topics/introduction/tutorials/igv-introduction/tutorial.html should probably be removed, that's not such a useful document / not meeting our style guide.

mvdbeek commented 3 years ago

jbrowse is fine for annotation, for variant analysis it's not so useful and IGV is the standard, would be a shame to lose this.

hexylena commented 3 years ago

Oh, no, I already updated that tutorial for jbrowse. Yeah, if igv doesn't work there, we just tell them to do jbrowse.

And yeah, it's a very fair point, not as functional as IGV for variant analysis (this is just for mapping though). But I get it. Unfortunately it's also a lot easier for people to run, even at the cost of functionality.