retrieve_SRR_fastq_files.qsub

#!/bin/bash -l
#$ -S /bin/bash
#$ -cwd
#$ -j y
#$ -m eas
#$ -N retrieve_SRR_fastq_files

# This script is a wrapper for the `fastq-dump` utility of the SRA Toolkit.
# It retrieves sequencing reads in FASTQ format (gzipped) from the NCBI
# Sequencing Read Archive (SRA) for a given set of SRR accession numbers,
# provided in a newline-delimited input file.  It creates files named with the
# SRR accession and the sequence (read/mate) number
# (e.g., `SRRnnnnnnnn_1.fastq.gz`, `SRRnnnnnnnn_2.fastq.gz`, etc.)

# It is intended for use with a cluster running Grid Engine and Environment
# Modules. It may also be run as a standalone bash script using bash instead of
# qsub.

# Adam Gower

# Parse command-line arguments
eval set -- "$(
  getopt --options=i:o: \
         --longoptions=input-file:,output-path: \
         --name "$0" -- "$@"
)"

while true
do
  case "$1" in
    -i|--input-file)
      inputFilename="$(readlink --canonicalize "$2")"
      shift 2 ;;
    -o|--output-path)
      outputPath="$(readlink --canonicalize "$2")"
      shift 2 ;;
    --)
      shift
      break ;;
    *)
      echo "Internal error"
      exit 1 ;;
  esac
done

if [[ "${inputFilename}" == "" || "${outputPath}" == "" ]]
then
  echo    "Usage:"
  echo    "  qsub [qsub flags] retrieve_SRR_fastq_files.qsub"
  echo    "       -i|--input-file [input filename]"
  echo    "       -o|--output-path [path to FASTQ files]"
  echo    "Options:"
  echo -n "  -i, --input-file      "
  echo    "Path to newline-separated file containing SRR accessions"
  echo -n "                        "
  echo    "(e.g., download from 'Accession List' in SRA Run Selector)"
  echo -n "  -o, --output-path     "
  echo    "Path where FASTQ files will be written "
  echo -n "                        "
  echo    "(will be created if it does not exist)"
else
  if [[ ! -e "${inputFilename}" ]]
  then
    # If the input file does not exist, print an error message and exit
    echo "File '${inputFilename}' does not exist; terminating."
  else
    # Load and list modules
    module load sratoolkit
    module list

    # Iterate over each SRR accession in the file, extracting FASTQ files
    # directly from SRA, with headers in original format
    # Note: the while loop construct is based on the Stack Overflow post:
    #       https://stackoverflow.com/questions/12916352#12919766
    #       It is written this way to handle input files that do not contain
    #       a terminal newline.
    while read accession || [ -n "${accession}" ]
    do
      echo "Extracting fastq.gz files from ${accession} to: '${outputPath}'"
      fastq-dump \
        --origfmt --split-files --gzip \
        --outdir "${outputPath}" \
        ${accession}
    done < "${inputFilename}"
  fi
fi