I want use deepbinner as the first step of a workflow. Deepbinner scans fast5 files for barcodes and then moves them to different folders accordingly. My problem is that files are sorted into a temp directory created by cwltool and disappear from their original location. The temporary files are then processed further and i get my desired output. However, all my raw data is lost.
Is there a way to prevent this behavior? I would like for the workflow to not alter the input directory at all, but i don't know how to make sure deepbinner only gets a copy of the directory.
Hello Tom,
Which CWL runner are you using, and which version is it?
Do your CWL descriptions use the
DockerRequirement
?Can you share a link to view your CWL descriptions?
I'm using cwltool version is 1.0.20181217162649 I am still struggling with getting my stuff to work with docker, so for now there is no DockerRequirement.
I added the relevant code to a repository: *edit: repository made private again.
Thank you for taking the time to help.
In case it is of interest: When running the tool in Docker it fails because it cannot move the input files out of their directories.
Using
InitialWorkDirRequirement
withwritable: true
solves the problem.