Hello,
I have a fastq that I'm trying to run cumulus/cellranger_workflow's cellranger count on. I've set up the directory structure as follows: gs://fc-e0000000-0000-0000-0000-000000000000/donor_1_fastq has 6 fastqs: donor_1_S1_L001_l1_001, donor_1_S1_L001_R1_001, donor_1_S1_L001_R2_001, donor_1_S1_L002_l1_001, donor_1_S1_L002_R1_001, and donor_1_S1_L002_R2_001. I believe this means that the donor's data is split between two samples(?), [donor_1_S1_L001_l1_001, donor_1_S1_L001_R1_001, donor_1_S1_L001_R2_001] and [donor_1_S1_L002_l1_001, donor_1_S1_L002_R1_001, donor_1_S1_L002_R2_001], and so I have placed each of these sets into subfolders, donor_1_S1_L001 and donor_1_S1_L002, respectively. The sample sheet, located alongside donor_1_fastq, then goes something like:
Sample,Reference,Flowcell,DataType
donor_1_S1_L001,GRCh38-2020-A,gs://fc-e0000000-0000-0000-0000-000000000000/donor_1_fastq,rna
donor_1_S1_L002,GRCh38-2020-A,gs://fc-e0000000-0000-0000-0000-000000000000/donor_1_fastq,rna
I believe I have set up the directory structure, named the files, and written the sample sheet as per the docs, yet I'm encountering the failure "Job cellranger_count.run_cellranger_count:NA:1 exited with return code 1 which has not been declared as a valid return code. See 'continueOnReturnCode' runtime attribute for more details." for each of the two samples. (Not sure where to find that attribute.)
I was hoping you could help me figure out what I'm setting up wrong. Maybe it's messing up because it's a single donor that for some reason is split in two, or something? Thanks,
Alexander Young