This guide outlines the process of downloading raw sequencing data from GEO/SRA using the NCBI SRA Toolkit. We will also cover troubleshooting common issues and tips for efficient downloading.

Prerequisites

1. Install the SRA Toolkit:
   • Download the latest version of the SRA Toolkit: NCBI SRA Toolkit.
   • Follow the installation instructions for your operating system.
2. Install Entrez Direct (optional):
   • Entrez Direct simplifies querying SRA databases. Installation guide: Entrez Direct.
3. Environment Setup:
   • Add the toolkit binaries to your PATH:
     bash

     Copy code

     export PATH=$PATH:/path/to/sratoolkit/bin

4. Verify Installation:
   • Run vdb-config -i to set up default download directories or change them to a directory with sufficient space.
   • Test the installation:
     bash

     Copy code

     prefetch --version
fastq-dump --version


Steps to Download Data

1. Identify the Accession Number

• Go to GEO: GEO
• Search for your dataset (e.g., GSE48215).
• Navigate to the sub-series or sample page (e.g., GSM1173000).
• Find the corresponding SRA links (e.g., SRP026538 or SRX317818).

2. Use prefetch to Download .sra Files

• Run the following command to download .sra files:
  bash

  Copy code

  prefetch -v SRR925811

  • By default, files are saved to /home/<USER>/ncbi/public/sra.

Note: If your home directory lacks space, update the cache location:

Change the “Workspace Name” to a directory on a larger disk.

3. Convert .sra to .fastq

• Use fastq-dump:
  bash

  Copy code

  fastq-dump --outdir /path/to/output/ --split-files /home/<USER>/ncbi/public/sra/SRR925811.sra

  • The --split-files option generates separate files for paired-end reads (e.g., _1.fastq and _2.fastq).

Example:

4. Download Multiple Runs

• Use esearch and efetch with xargs for batch downloads:
  bash

  Copy code

  esearch -db sra -query "PRJNA40075" | efetch --format runinfo | cut -d ',' -f 1 | grep SRR | xargs -I {} prefetch {}

• Convert all downloaded .sra files to .fastq:
  bash

  Copy code

  for file in /path/to/sra/*.sra; do
fastq-dump --split-files "$file"
done


5. Using Docker for SRA Toolkit

• Use a pre-built Docker container:
  bash

  Copy code

  docker run -v $PWD:$PWD quay.io/biocontainers/sra-tools:3.1.1--h4304569_2 prefetch SRR28520745 --output-directory $PWD/
docker run -v $PWD:$PWD quay.io/biocontainers/sra-tools:3.1.1--h4304569_2 fastq-dump --outdir $PWD/SRR28520745/ --split-files $PWD/SRR28520745/SRR28520745.sra


6. Alternative: Direct Download from ENA

• If you encounter issues with SRA Toolkit, download directly from ENA:
  • Visit: EBI ENA
  • Search for the dataset (e.g., SRX317818).
  • Download .fastq files via FTP:
    bash

    Copy code

    wget ftp://ftp.sra.ebi.ac.uk/vol1/fastq/SRR925/SRR925811/SRR925811_1.fastq.gz
wget ftp://ftp.sra.ebi.ac.uk/vol1/fastq/SRR925/SRR925811/SRR925811_2.fastq.gz


7. Troubleshooting Common Errors

• UTF-8 Character Error:
  • Ensure no special characters in file paths. Re-run:
    bash

    Copy code

    fastq-dump --outdir /clean/path/ --split-files /path/to/SRR6294675.sra

• No Accession to Process:
  • Check the .sra file exists and paths are correct.
• Command Not Found:
  • Verify PATH includes the SRA Toolkit binaries.

8. Automate Download with a Script

9. Further Processing

• Align .fastq files using an aligner (e.g., STAR, BWA, TopHat).
• Perform variant calling or downstream analysis.

10. Recommended Reading

• SRA Toolkit Documentation
• Galaxy RNA-Seq Tutorial

By following this guide, you can efficiently download and prepare sequencing data for downstream analysis.