Skip to content

Latest commit

 

History

History
93 lines (63 loc) · 3.32 KB

how-to_SRA-submission.md

File metadata and controls

93 lines (63 loc) · 3.32 KB
Raw

How to submit to the NCBI's SRA?

author: Timothée Flutre (INRA)

version: 06/09/2018

Goal: insert sequencing reads to the SRA database of the NCBI.

All relevant, up-to-date information is in the official doc: this is only a quick tutorial!

Prerequisites

Raw data

List all your fastq files (compressed with gzip) in a single file:

mkdir upload_NCBI
cd upload_NCBI
ls /my/data/*.fastq.gz > list_files_NCBI-SRA.txt

Number of files:

cat list_files_NCBI-SRA.txt | wc -l

Number of samples:

cat list_files_NCBI-SRA.txt | awk '{split($0, a, "_R"); print a[1]}' | sort | uniq -c | wc -l

Meta-data

Make sure you have the passport data of each sample: unique identifier/code of the sample or the genotype, unique identifier/code of the cultivar, the sample names, species, etc.

For grapevine at INRA, we should use the database of the RFCV.

Create an account

Go here.

Create a BioProject submission

Go here and follow the steps.

Example:

  • enter your Project title (e.g. "RAD-seq of ABC");
  • enter your Public description;
  • enter your Relevance (e.g. "agricultural");
  • enter your Grants (e.g. "2014-ABC ; CASDAR");
  • enter your Biosample type (e.g. "Plant");
  • skip BioSamples for the moment;
  • finish.

The status of your submission is available in your submission portal. Once processed, retrieve there your BioProject identifier, e.g PRJNA725341.

Create a BioSample submission

Go here and follow the steps.

Advice: download the .tsv file and fill it programatically with your favorite programming language (R, Python, Perl, etc).

Notes:

  • one can't submit a file with more than 1000 samples;
  • the file should be in ASCII;
  • one can't submit a sample already existing in the NCBI (the submission interface automatically detects such cases).

The status of your submission is available in your submission portal. Once processed, download the attributes.tsv file with BioSample accessions.

Create a SRA submission

Go here and follow the steps.

Download the .tsv file and fill it programatically with your favorite programming language (R, Python, Perl, etc).

Notes:

  • make symbolic links of all fastq files into the same directory to ease the upload;
  • use the Aspera Command-Line tool (FASP) instead of FTP to upload many samples because it is much faster.
ls fastq_gbs-279/* | wc -l
ls fastq_gbs-279/*_R*.fastq.gz | wc -l
ls -la fastq_gbs-279/ | grep "\->" | wc -l
nohup ascp -i ~/.ssh/aspera.openssh -QT -l100m -k1 \
      -d fastq_gbs-279 \
      [email protected]:uploads/your@email_<...> \
      >& stdout_ascp.txt &