Working with NCBI SRA files

Working with SRA files can be tricky sometimes. Here I have short description about how to download and process SRA files.

NGS read files can be downloaded using sra-toolkit. First you have to download latest version of sra-toolkit from their website. The file should be in tar.gz format.

Then untar the file based on the OS you are using.

Then use the following command to put command line in path

export PATH=$PATH:/home/joshil/ngs/rnaseq/sratoolkit.2.9.0-centos_linux64/bin

Remember to change the path. Here I have used my path as an example (/home/joshi/ngs/rnaseq)

Then navigate to sratoolkit.2.9.9-ubuntu/bin folder. Your version might be different. Inside the bin folder you will be able to see all the executables. You can use each script as shown below

Eg.

./fastq-dump
./prefetch 

Suppose you want to download a SRA file named SRR6047337. Then simply use follwing command

./prefetch SRR6047337

This will download SRR6047337.sra files. Don’t see the file ? Well, here is the trick. All the files are downloaded in home/ncbi/public/sra folder. This prefetch command automatically creates a ncbi folder under the home folder.

Now, you need fastq files (may be paired or single) instead of .sra files. Use following command to convert .sra file to the fastq file.

./fastq-dump --split-3 SRR6047337 

It will take specified file from the $home/ncbi/public/sra folder and converts it to SRR6047337_1.fastq and SRR6047337_2.fastq.

Hope this will help you to download SRA files from NCBI.

Image credit: Unsplash