10x genomics single cell RNA-Seq data submission to GEO

Question

[Nitin Narwade]

Hi,

I have a very basic question about submitting scRNA-Seq data (10X genomics) to GEO. Do I need to submit the demultiplexed raw data or just the sample-wise data submission is enough?

Any help or suggestions would be really appreciated.

Thanks in advance.

Answer 1

[benformatics]

You could submit the raw data (i.e. 2-4 fastqs) if you wanted and from some perspectives, for instance raw "reproducibility", that might be ideal. However, a lot of labs seems to push the processed BAM files produced from the Space Ranger/Cell Ranger pipeline which include a lot of the important information (e.g. cell barcodes/UMI) to the SRA which is then linked to the project on GEO and within that project they include additional processed data from the pipeline (e.g. gene by cell count matrix, whitelist of barcodes per sample, etc...).

For downstream external research, I find the pulling the processed BAM's ideal because then I know that when I work with the data I'm using the same exact alignment parameters and aligner/cell ranger software versions that you used during your experiment.

Re-reading your question: You absolutely should be pushing some level of raw nucleic acid information either FASTQs or BAMs with your submission to the GEO (ultimately they would appear on the SRA unless they need to protected for privacy issues). Processed data goes on the GEO and raw data nt data goes to the SRA.

[Nitin Narwade]

Perfect, thank you so much for the information :)