I want to do assembly and quantification for differential gene expression of my datasets. I use StringTie to run the bam files from STAR. But it’s been driving me nuts… Because, it worked well on some datasets and always give errors (Fatal error: Exit code 1 ()) on some others. Although I load the all together, so same parameters.
When I re-run StringTie, the success run and error is also not always not the same datasets. At first I report the error and Galaxy EU upgraded the version. It’s looks like the problem is still there…
Hi @zee_azizah
do you submit StringTie jobs on a collection? Some jobs might fail sporadically when submitted on a collection. If yes, re-run the failed jobs and activate ‘Attempt to re-use jobs with identical parameters?’ at the bottom of the job setup page. If I remember correctly, the completed job will replace the failed job. Alternatively, submit jobs on individual datasets. In my experience, jobs on individual datasets are less prone to failures. I have not done any stats, but it looks like this.
Kind regards,
Igor
Hi @igor
I did not submit the job as collection. But still gave a random failed jobs. When I re-run the failed datasets one at a time, sometimes it ran successfully sometimes it still give random error.
Hi @zee_azizah
I hope Galaxy Europe admin can comment on this. It might be a server related issue: jobs go to different nodes, and the observed pattern can be explained by a faulty node. You can report the error for one job: click on any output of the failed job, click at error icon, the one looking like lady bird bug/beetle, add the description, just like you did here, and hit Report. The report goes to the support team, so they can check what is going on.
Kind regards,
Igor
