Strand Bias Analysis Tool

Overview

SBAT is a Python command-line tool for detection of strand bias. Strand bias is a situation when information from one strand of DNA is overrepresented compared to the information from the other strand. It is one of the types of bias that occur in next-generation sequencing data. Strand bias might lead to incorrect evaluation of results gained from sequencing data, if the bias is high. This tool offers a way of validating quality of the data against strand bias. More about strand bias and development of this tool can be found [here](path to bachelor thesis once it is public).

The tool uses Jellyfish k-mer counting tool for counting k-mers in the NGS data and compares frequencies of k-mers and their complements, creating both statistics and visual analysis of the results of strand bias.

Installation

First, Jellyfish must be installed.

On Debian and Ubuntu with apt:

sudo apt update
sudo apt install jellyfish

On MacOS with brew:

brew install jellyfish

On Arch, it is available from AUR.

On Windows, the best option is to use WSL. For other OS or installation from source code, see here

After Jellyfish is installed, proceed with SBAT itself:

Installation from pip

pip install sbat

To install from source code, download the code and run following in the root of the source tree:

python3 -m pip install --upgrade build
python3 -m build
pip install -e .

Usage

In order to perform analysis on one or multiple files, use command sbat followed by your files. By default, k-mers for k in range 5 to 10 are analysed. The figures and statistics of the tool are saved into sbat_out directory, which will be created if does not exist already. By default all the partial results of the SBAT (Jellyfish-generated files, statistics for each size of k) are deleted at the end of the run. To prevent that, use argument -c

sbat my_file.fasta my_file2.fasta my_file3.fastq

Following command additionally specifies output directory with -o and keeps partial results of computations using parameter -c. To speed up SBAT run time, use parameter -t T with specified number of threads you wish to pass to the application. To specify size of k-mers for which you want to run analyses, use parameter -m START END. If one argument is passed to it, SBAT runs only for this size of k. If two arguments are passed, application analyses k-mers in range [START, END]

sbat my_file.fasta my_file2.fasta my_file3.fastq -o output_dir -c -t 10 -m 5 8

If you want to analyse Nanopore dataset, add -n in order to run more specific, time-based analysis. As part of this analysis, dataset is divided into one-hour long bins. Each of them is then analysed on its own. The time duration of one bin can be set by -i H parameter followed by number of hours. If you wish to subsample your data, you can use parameters -r N or -b N to take only first N reads or bases of each bin.

sbat my_nanopore.fastq -o output_dir -b 500M -i 4 -n

To see all possible options, run:

sbat -h

From version 0.0.9, -p parameter enables creation of interactive plots as well as .jpg results. After analysis finishes, SBAT creates Bokeh server on http://localhost:5006/