Tutorial:How to upload your RNA-Seq data to NCBI Sequence Read Archive (SRA)

Question

[CandiceChuDVM]

Hi all,

I am writing this tutorial in response to my previous question: NCBI SRA submission: neither sample_name nor biosample_accession are set.

I took me three days to figure our what's going on so I hope my tutorial can save your time and make your life easier:

Tutorial: How to upload your data to the evil Sequence Read Archive (SRA)?

Enjoy.

[Istvan Albert]

Thanks for putting this all together.

Right from the start, just getting data in and out of the SRA demonstrates the complete disconnect between the theory and the reality of reproducible research.

[CandiceChuDVM]

I can't agree more with you.

[igor]

Not sure I understand this part:

My files are stored on supercomputer and I am not going to download them to local computer then upload them through FTP.

As you mention elsewhere, it's running linux. In that case, it should have FTP. You can use which ftp to confirm.

[CandiceChuDVM]

Yes. I have tried ftp using the following instruction:

After you login,
1. navigate to your account folder: cd uploads/[e-mail]_TsOpWGZR
2. create a subfolder with a meaningful name (required): mkdir new_folder
3. navigate to that folder: cd new_folder
4. deposit your files into that folder: put file_name

However, I found that if I am not going to write a script for the process, it's easy for me to make mistake and get kicked out from the log in status. That's why I choose to use Aspera in which I can only put one line of command to achieve what I want.

[Lucila]

Hi CandiceChuDVM,

thank you so much for this post. It will be very useful.

I would like to ask you one question: Do you know if they have requirements to let you upload you data (in terms of quality and contamination of your samples, for example with adaptors)?

Thank you!

Cheers, Lucila.

[igor]

I don't think they check the contents of your files other than to make sure they are in the proper format.

Answer 1

[igor]

According to NCBI, you should be submitting RNA-seq data to GEO, not SRA:

Functional genomics studies that examine gene expression, regulation or epigenomics (using methods such as RNA-Seq, miRNA-Seq, ChIP-Seq or methyl-Seq) should be submitted to GEO.

Source: https://www.ncbi.nlm.nih.gov/guide/howto/submit-sequence-data/

As an added benefit, it's a much simpler process.

[GenoMax]

GEO submits raw data to SRA on your behalf automatically (https://www.ncbi.nlm.nih.gov/geo/info/seq.html)

[igor]

Yes, but that part is automated, so you don't have to worry about that as a submitter.

[CandiceChuDVM]

Thanks! I wasn't aware of that. I will contact SRA later to figure out what to do at this point.

[igor]

If you email GEO, they can link SRA and GEO entries. Or at least they used to.