Cell Ranger ARC null path value

[ChristopherBarrington]

Description of the bug

Similar to #374 perhaps, I get a Path value cannot be null error when running --aligner cellrangerarc.

My sample sheet is:

sample,fastq_1,fastq_2,fastq_barcode,sample_type
NestinGFP8w_3,inputs/fastq/20241018_LH00442_0061_A22NY2YLT3/fastq/AHM4688A49_S204_L008_R1_001.fastq.gz,inputs/fastq/20241018_LH00442_0061_A22NY2YLT3/fastq/AHM4688A49_S204_L008_R2_001.fastq.gz,,gex
NestinGFP8w_3,inputs/fastq/20241025_LH00442_0062_A22NTH3LT3/fastq/AHM4688A49_S22_L005_R1_001.fastq.gz,inputs/fastq/20241025_LH00442_0062_A22NTH3LT3/fastq/AHM4688A49_S22_L005_R2_001.fastq.gz,,gex
NestinGFP8w_3,inputs/fastq/20241008_LH00442_0059_A223NY3LT1/fastq/AHM4688A50_S3_L001_R1_001.fastq.gz,inputs/fastq/20241008_LH00442_0059_A223NY3LT1/fastq/AHM4688A50_S3_L001_R2_001.fastq.gz,inputs/fastq/20241008_LH00442_0059_A223NY3LT1/fastq/AHM4688A50_S3_L001_I2_001.fastq.gz,atac
NestinGFP8w_3,inputs/fastq/20241008_LH00442_0059_A223NY3LT1/fastq/AHM4688A50_S3_L002_R1_001.fastq.gz,inputs/fastq/20241008_LH00442_0059_A223NY3LT1/fastq/AHM4688A50_S3_L002_R2_001.fastq.gz,inputs/fastq/20241008_LH00442_0059_A223NY3LT1/fastq/AHM4688A50_S3_L002_I2_001.fastq.gz,atac

(This confused me as well because the sample_type:gex has included atac libraries too?)

An example of an entry in ch_fastq is below:

[DUMP] [
    {
        "id": "NestinGFP8w_3",
        "expected_cells": "",
        "seq_center": "",
        "sample_type": "gex",
        "feature_type": "",
        "single_end": "false"
    },
    [
        "/inputs/fastq/20241018_LH00442_0061_A22NY2YLT3/fastq/AHM4688A49_S204_L008_R1_001.fastq.gz",
        "/inputs/fastq/20241018_LH00442_0061_A22NY2YLT3/fastq/AHM4688A49_S204_L008_R2_001.fastq.gz",
        "/inputs/fastq/20241025_LH00442_0062_A22NTH3LT3/fastq/AHM4688A49_S22_L005_R1_001.fastq.gz",
        "/inputs/fastq/20241025_LH00442_0062_A22NTH3LT3/fastq/AHM4688A49_S22_L005_R2_001.fastq.gz",
        "/inputs/fastq/20241008_LH00442_0059_A223NY3LT1/fastq/AHM4688A50_S3_L001_R1_001.fastq.gz",
        "/inputs/fastq/20241008_LH00442_0059_A223NY3LT1/fastq/AHM4688A50_S3_L001_R2_001.fastq.gz",
        "/inputs/fastq/20241008_LH00442_0059_A223NY3LT1/fastq/AHM4688A50_S3_L002_R1_001.fastq.gz",
        "/inputs/fastq/20241008_LH00442_0059_A223NY3LT1/fastq/AHM4688A50_S3_L002_R2_001.fastq.gz"
    ]
]

Since (I think) my channel has too many files and a structure that is not expected by the CELLRANGERARC_COUNT module, I suspect I have misconfigured something?

Sorry for being so vague, I am at a loss of what to try next.

Command used and terminal output

nextflow run nf-core/scrnaseq \
-revision 2.7.1 \
-config custom.config \
-profile crick \
-resume \
--input $sample_sheet \
--outdir outputs \
--aligner cellrangerarc \
--cellranger_index $index

-[nf-core/scrnaseq] Pipeline completed with errors-
WARN: Input tuple does not match tuple declaration in process `NFCORE_SCRNASEQ:SCRNASEQ:CELLRANGERARC_ALIGN:CELLRANGERARC_COUNT` -- offending value: [[id:NestinGFP8w_3, expected_cells:null, seq_center:null, sample_type:gex, feature_type:null, single_end:false], [/inputs/fastq/20241018_LH00442_0061_A22NY2YLT3/fastq/AHM4688A49_S204_L008_R1_001.fastq.gz, /inputs/fastq/20241018_LH00442_0061_A22NY2YLT3/fastq/AHM4688A49_S204_L008_R2_001.fastq.gz, /inputs/fastq/20241025_LH00442_0062_A22NTH3LT3/fastq/AHM4688A49_S22_L005_R1_001.fastq.gz, /inputs/fastq/20241025_LH00442_0062_A22NTH3LT3/fastq/AHM4688A49_S22_L005_R2_001.fastq.gz, /inputs/fastq/20241008_LH00442_0059_A223NY3LT1/fastq/AHM4688A50_S3_L001_R1_001.fastq.gz, /inputs/fastq/20241008_LH00442_0059_A223NY3LT1/fastq/AHM4688A50_S3_L001_R2_001.fastq.gz, /inputs/fastq/20241008_LH00442_0059_A223NY3LT1/fastq/AHM4688A50_S3_L002_R1_001.fastq.gz, /inputs/fastq/20241008_LH00442_0059_A223NY3LT1/fastq/AHM4688A50_S3_L002_R2_001.fastq.gz]]
ERROR ~ Error executing process > 'NFCORE_SCRNASEQ:SCRNASEQ:CELLRANGERARC_ALIGN:CELLRANGERARC_COUNT (1)'

Caused by:
  Path value cannot be null

Relevant files

No response

System information

• Nextflow/24.04.2
• Singularity/3.11.3

[RPSeaman]

I also just came across this issue, except of the FASTQC step. Couple of additional pieces that I saw that may potentially be helpful:

This warning produced almost immediately upon launch

WARN: Samplesheet warnings:
	The samplesheet contains following unchecked field(s): [fastq_barcode]

And the error which looks the same except on FASTQC instead of CELLRANGERARC_COUNT

ERROR ~ Error executing process > 'NFCORE_SCRNASEQ:SCRNASEQ:FASTQC_CHECK:FASTQC (4)'

Caused by:
  Path value cannot be null

My sample sheet looks the same except for my read naming is the alternative option for atac:

sample,R1,R3,R2,atac

instead of

sample,R1,R2,I2,atacf

Thanks.

[elanstop]

I am having the same problem. This is my samplesheet (following the guidelines in the docs):

sample,fastq_1,fastq_2,fastq_barcode,sample_type
downsampled10kpbmc,/home/jovyan/work/workbench/multiome_data/downsampled10kpbmc_atac/downsampled10kpbmc_atac_S1_L001_R1_001.fastq.gz,/home/jovyan/work/workbench/multiome_data/downsampled10kpbmc_atac/downsampled_10kpbmc_atac_S1_L001_R3_001.fastq.gz,/home/jovyan/work/workbench/multiome_data/downsampled10kpbmc_atac/downsampled_10kpbmc_atac_S1_L001_R2_001.fastq.gz,atac
downsampled10kpbmc,/home/jovyan/work/workbench/multiome_data/downsampled10kpbmc_gex/downsampled10kpbmc_gex_S1_L001_R1_001.fastq.gz,/home/jovyan/work/workbench/multiome_data/downsampled10kpbmc_gex/downsampled10kpbmc_gex_S1_L001_R2_001.fastq.gz,,gex

This is the command I'm running:

nextflow run main.nf -profile docker --input /home/jovyan/work/workbench/elan_scrnaseq_test_samplesheet.csv --outdir /home/jovyan/work/workbench/scrnaseq --aligner cellrangerarc --skip_fastqc --skip_multiqc --cellranger_index /home/jovyan/work/workbench/multiome_data/refdata-cellranger-arc-GRCh38-2020-A-2.0.0

And this is the error:

WARN: Input tuple does not match tuple declaration in process `NFCORE_SCRNASEQ:SCRNASEQ:CELLRANGERARC_ALIGN:CELLRANGERARC_COUNT` -- offending value: [[id:downsampled10kpbmc, expected_cells:[], seq_center:[], sample_type:atac, feature_type:[], single_end:false], [/home/jovyan/work/workbench/multiome_data/downsampled10kpbmc_atac/downsampled10kpbmc_atac_S1_L001_R1_001.fastq.gz, /home/jovyan/work/workbench/multiome_data/downsampled10kpbmc_atac/downsampled_10kpbmc_atac_S1_L001_R3_001.fastq.gz, /home/jovyan/work/workbench/multiome_data/downsampled10kpbmc_gex/downsampled10kpbmc_gex_S1_L001_R1_001.fastq.gz, /home/jovyan/work/workbench/multiome_data/downsampled10kpbmc_gex/downsampled10kpbmc_gex_S1_L001_R2_001.fastq.gz]]
ERROR ~ Error executing process > 'NFCORE_SCRNASEQ:SCRNASEQ:CELLRANGERARC_ALIGN:CELLRANGERARC_COUNT (1)'

Caused by:
  Path value cannot be null

Earlier in the stdout I also got a warning about fastq_barcode being an unidentified header. I don't understand how this is possible given that that header is included in the example template. Overall it seems that either the example samplesheet from the docs is incorrect or the parsing of the samplesheet is going wrong

[sbgonenc]

I am also having exactly the same problem. I would be glad to hear if there is a work around or any updates?