Hello! I posted this in #CHN5BV5DW but got no replies yet. I have been trying to use scrnaseq to map CITE-Seq samples. The pipe always breaks on the antibody file mapping phase. I get the same error if I use cellranger count or multi. But I discovered that I can run this with just with cellranger count (no nf-core pipeline). All the cellranger software needs is library_type info: Antibody Capture/Gene Expression This is the beginning part of scrnaseq pipe error:

ERROR ~ Error executing process > 'NFCORE_SCRNASEQ:SCRNASEQ:CELLRANGER_ALIGN:CELLRANGER_COUNT

Process NFCORE_SCRNASEQ:SCRNASEQ:CELLRANGER_ALIGN:CELLRANGER_COUNT (sample_1_FB) terminated with an error exit status (1)

2024-09-05 11:57:34 [runtime] (failed)          ID.<http://sample_1_FB.SC|sample_1_FB.SC>_<http://RNA_COUNTER_CS.SC|RNA_COUNTER_CS.SC>_MULTI_CORE.MULTI_CHEMISTRY_DETECTOR.DETECT_COUNT_CHEMISTRY

[error] Pipestance failed. Error log at:

sample_1_FB/SC_RNA_COUNTER_CS/SC_MULTI_CORE/MULTI_CHEMISTRY_DETECTOR/DETECT_COUNT_CHEMISTRY/fork0/chnk0-u004ed98050/_errors

It goes on forever but I think the issue is that the nfcore's scrnaseq pipe is not transmitting the info about GEX or AB from the sample table to the cellranger properly. just for clarity here's the command I run:

nextflow run nf-core/scrnaseq -r 2.7.1 -profile cluster --aligner cellranger --skip_fastqc true --input /location/cellranger_sample_sheet.csv

--fasta /location/fasta/genome.fa --gtf /location/genes/genes.gtf --fb_reference /location/CITE_seq_barcodes.csv --outdir /location/01_mapped/

-c /location/nextflow_sc_extraresource.config -w /location/username/work/ --protocol auto

I have tried specifying feature_type as gex or ab but I still get the same error. So the long story short, the main issue is that I don't know how to properly configure scrnaseq pipeline for cite-seq samples or even if that can be done with the current pipeline.

Have a look at the cellranger command in

.command.sh

in the task working directory and compare it to the command line that you are using. If something is missing, you might be able to provide it to the process via

ext.args

in a custom config file.

I didn't just have one .command.sh file but several dozen and all of them in the -w work folder I specifyed. Most of them say: Automatically rename staged files for input into cellranger count. followed by lots of renaming code and the rest were: cellranger mkgtf followed by a few lines of code. So it now looks like the pipe doesn't get past make gtf stage or crashes just after that. In the overview panel on command line it looks like it always crashes during the count phase. I read through the scrnaseq user page several times and tried all the options to specify additional file type info. There's nothing for cellranger count but there is one for cellranger multi and one for cellranger arc. If I specify for the multi it still doesn't work. If I use the same way to specify for cellranger count it still doesn't work and gives the same error(s).