cellranger count for multiple samples

Question

[Assa Yeroslaviz]

I was wondering if it is possible to run cellranger count with multiple samples in one go and get an output folder for each sample separately.

I'm using the following command to run cellranger count

cellranger count --id=10X_V211 \
           --libraries=library.csv \
           --transcriptome=Drerio_genome/ \
           --create-bam=true \
           --output-dir v211_Drerio/

while my library.csv file is this

fastqs,sample,library_type,
fastq_files/,low1,Gene Expression,
fastq_files/,low2,Gene Expression,
fastq_files/,high3,Gene Expression,
fastq_file/,high4,Gene Expression,

But this would treat all the samples as just one and would create just one output file.

Is there a way to handle those samples separately or do I need to run the command for times and use the --fastq and --sample parameters instead?

thanks

Assa

Answer 1

[Dave Carlson]

There are plenty of ways that you could automate the process of running Cellranger across multiple samples.

I would recommend checking out the nf-core scrnaseq pipeline. It's designed for situations like this.

[ATpoint]

Why would you run CellRanger, which is fully self-contained and automated additionally via something as heavy as an nf-core pipeline?

[Dave Carlson]

Why not? It runs some QC steps beforehand, produces a nice MultiQC report that summarizes the results for each of the sample together, and it has the option of running alternatives to Cellranger as well (if that's of any interest).

The OP's goal appears to be to parallelize (or at least automate) the workflow across multiple samples. This is one way to do that.

I guess I run nf-core pipelines so often that they don't seem particularly "heavy" to me either.