cellranger multi => pipeline creates STAR index that does not match probeset

[nick-youngblut]

Description of the bug

If the user provides --fasta and --gtf instead of --cellranger_index, then the pipeline will created a STAR index for the cellranger multi job. However, this index does not match the probeset and results in the error:

Reference genome does not match probe set CSV reference "GRCh38"

The pipeline also removes the comment lines from the GTF:

##description: evidence-based annotation of the human genome (GRCh38), version 32 (Ensembl 98)
##provider: GENCODE
##contact: [email protected]
##format: gtf
##date: 2019-09-05

...however, that doesn't seem to matter for cellranger multi.

So, either the user should not be able to use --fasta and --gtf instead of --cellranger_index when running the cellranger multi workflow in scrnaseq, or the creation of the STAR index must be modified so that the index matches the probeset.

Command used and terminal output

nextflow run main.nf
--aligner cellrangermulti
--skip_cellrangermulti_vdjref
--gex_frna_probe_set Chromium_Human_Transcriptome_Probe_Set_v1.0.1_GRCh38-2020-A.csv
--cellranger_multi_barcodes sample_barcodes.csv
--cellranger_index refdata-gex-GRCh38-2020-A/
--input samples.csv
--outdir scrnaseq_output

Relevant files

No response

System information

• nextflow: 24.04.4
• hardware: HPC
• executor: SLURM
• container engine: Apptainer
• os: linux
• scrnaseq: v2.7.1