how to combine multiple Cellranger count results from a single library

Question

[g744695539]

Hi All,

I have many fastq files from the same library. If I run cellranger count on all of them together, I can only do it through slurm by submitting to a single node. I want to use multiple nodes, so that I can get multi cellranger count results. In this case, how can I combine all the cellranger count results together in the end?

Answer 1

[LChart]

I have many fastq files from the same library. If I run cellranger count on all of them together, I can only do it through slurm by submitting to a single node. I want to use multiple nodes, so that I can get multi cellranger count results

Why do you want multiple count results only to combine them in the end? Strictly speaking, you cannot do this. Because you've sequenced the same library multiple times, then if you process those sequencing runs separately, you will vastly overcount fragments. This is because duplicate molecules (the same cell barcode, UMI, and fragment) will be present in each of the fastQ files, but can only be de-duplicated within individual fastQ files.

The proper way to do this is to use ln -s to softlink (or simply copy) all of the fastQ files into the same directory, in the form samname_S001_L00{1,2,3,4,etc}_....fastq.gz and run cellranger appropriately.

[g744695539]

Thank you very much! Because I don't know how to run cellranger count across multiple nodes via SLURM. I can only run cellranger count on a portion of the FASTQ files on node 1, and another portion on node 2.