Hi all,
I have been trying to work through the training tutorial for hybrid assembly using unicycler, and the tool continues to fail using both the tutorial data and personal data: Unicycler Assembly
The error(s) displayed:
WARNING:galaxy.model:Datatype class not found for extension ‘gfa1’
Error: SPAdes encountered an error:
Traceback (most recent call last):
File “/dep/_conda/envs/mulled-v1-db3e342689a12eb9f9b15f8e2f2495f6124f67ba6fc02c95955211476a2b12c5/bin/spades.py”, line 659, in
main(sys.argv)
Which public server where you working at: server URL or describe if other (local?)
Copy and paste back the Dataset details page (middle part of the view) with the inputs expanded to show the peek view. Or, a share link to the entire history for the quickest help from more people.
I am currently working on the FDA maintained Galaxy Server Galaxytrakr.org, but I also tried the standard US instance, UseGalaxy.org and the same problems occurred.
Update: The tests at the GalaxyTrakr server failed. Bug reporting is disabled at the server … but you could send them an email and describe what is going on. This is something only the administrators can fix up.
The best advice is to try at a different server. See the “available at these Galaxies” menu in the top information box of the tutorial. If listed, the UseGalaxy.* servers are top choices.
The tutorial inputs look OK from a quick look. The issues seems to be technical and specific to the GalaxyTrakr server. I wasn’t able to reproduce the problem at UseGalaxy.org but you could post back your history of a failed test run there, too, and I’ll take a look.
There are two related things going on at GalaxyTrakr
The server is running an earlier version of Galaxy (21.05 according the to Help menu there). Galaxy is at version 23.0 now.
The version of Unicycler executed was the most current version installed on the server. They also host an earlier version.
What I am guessing is going on → That wrapper was recently updated and tested against the latest version of Galaxy (23.0). Why? The expanded stderr ends with a message about a technical item the tool wrapper calls, but the server couldn’t find/execute that bit. That seems related to how the optional graph outputs were posted back to the history – the job itself seems completed.
What to try
Start over in a new history, give it a distinct name, then load the data again. (this is just to get rid of any potential hiccups). Using all defaults with the Upload tool usually works best, meaning let Galaxy autodetect/guess datatypes. You can adjust after if the miscall was something simple (fastq → fastqsanger). If totally wrong or the upload isn’t complete, then skip the rest and report this problem first.
Try a rerun with the original version of the tool and choose the first output option in the menu (assembly only, no graphs).
To compare, I started tests in my copy of your history, and those are still running. Everything is tagged. Galaxy | GalaxyTrakr 🧬🔬. If the job fails, I’ll submit a bug report.
You should do the same.
If the job errors again, then you should submit a bug report to the server admins. In the comments you can include the tutorial link and a link to this topic – both for extra context. In addition, you could send them an email with the same information plus the shared history link. They’ll be able to review and sort out what is going on.
view) with the inputs expanded to show the peek view. Or, a share link to the entire history for the quickest help from more people.