The problem with scatter with the SRA data

[spcspin]

We try to automatically process multiple sra files by using cwl's scatter function. We ran it successfully with yeast data, but when we switched to bee data, we kept encountering slurm restarting the whole job in the middle of the run, so we couldn't get any results.

[mpoelchau]

Another problem we observed is that the output files will only be available at the end of the scatter workflow. If the workflow fails after several accessions completed, none of the results are available.

[spcspin]

The scatter.cwl was tested on two bee files and completed successfully. However, it is processed sequentially, not parallelly.

[spcspin]

Another option: using loop