srautils

srautils is a program used for download and dump NCBI SRA archive raw fastq data. It provides a fast and easy way to fetch sra data and covert sra file into fastq/fasta sequence data for our scientific research.

1. Requirement

• Linux64
• python >=3.8
• sratookits

2. Install

The latest release can be installed with

pypi:

pip3 install srautils -U

The development version can be installed with：(for recommend)

pip3 install git+https://github.com/yodeng/srautils.git

3. Usage

srautils include srautils fetch and srautils dump sub-commands.

3.1 srautils fetch

The fetch command is used for download SRA file by only giving an accession SRA id, it's a rapid and interruptable download accelerator. All original SRA files are obtained directly from AWS Cloud with UNSIGNED assess. This tools split the hold download into many pieces and record the progress for each chunk in a *.ht binary file, this can significantly speed up the download. Auto resume can be running by loading the progress file if any interruption. Command help as follows:

$ srautils fetch -h 
usage: srautils fetch [-h] -i <str> [-o <str>] [-n <int>] [-s <str>]

optional arguments:
  -h, --help            show this help message and exit
  -i <str>, --id <str>  input sra-id, SRR/ERR/DRR allowed, required
  -o <str>, --outdir <str>
                        output sra directory, current dir by default
  -n <int>, --num <int>
                        the max number of concurrency, default: auto
  -s <str>, --max-speed <str>
                        specify maximum speed per second, case-insensitive unit support (K[b], M[b]...), no-limited by default

3.2 srautils dump

The dump command is a parallel fastq-dump wrapper which used for dump SRA file and get the raw fastq/fasta sequence data as output.

NCBI fastq-dump is very slow, even if you have high machine resources (network, IO, CPU). This tool speeds up the process by dividing the work into multiple jobs and runing all chunked jobs parallelly in localhost or sge cluster (default) environment. After chunk jobs finished, all resuslts will be concatenate together. The command usage below here:

$ srautils dump -h 
usage: srautils dump [-h] -i <file> [-o <dir>] [-p <int>] [-q [<str> ...]] [-l <file>] [--no-gzip] [--fasta] [--local]

optional arguments:
  -h, --help            show this help message and exit
  -i <file>, --input <file>
                        input sra file, required
  -o <dir>, --outdir <dir>
                        output directory, current dir by default
  -p <int>, --processes <int>
                        number of dumps processors, 10 by default
  -q [<str> ...], --queue [<str> ...]
                        sge queue, multi-queue can be sepreated by whitespace, all.q by default
  -l <file>, --log <file>
                        append srautils log info to file, stdout by default
  --no-gzip             do not compress output
  --fasta               fasta only
  --local               run sra-dumps in localhost instead of sge

4. License

srautils is distributed under the MIT licence.

5. Reference

• NIH NCBI Sequence Read Archive (SRA) on AWS
• ncbi/sra-tools