Quick tutorial (installation, create index, and first run)

Installation

1. Install cwltool and python packages.

If you are not using python3, install python3 via anaconda or another way. Then, install these;

pip install cwlref-runner numpy pandas scipy statsmodels seaborn

2. Install Docker

Install docker (alternatively, you can use udocker when you have no root privileges. See Tips section.)

3. clone the repo and add the path.

git clone https://github.com/yyoshiaki/VIRTUS

You can add ./VIRTUS/bin to PATH in .zshrc or .bashrc etc.

Create indices

just run the code below. A set of indices will be created in your current directory. You can specify a yaml file, options or just a url as the input.

createindex.cwl https://raw.githubusercontent.com/yyoshiaki/VIRTUS/master/workflow/createindex.job.yaml

First run

VIRTUS has a great wrapper for multiple samples. The input is a comma-separated text file or CSV file. The first column is arbitral sample names, the second is SRR id, or fastq files (when you specify --fastq option). Note that --fastq requires the suffix removed file names. Refer to the documentation in more detail. The third column is the sequence layout (SE or PE), and the Fourth is groups. Let's create an example csv file (or download it).

input.csv

Name,SRR,Layout,Group
Flu_1,SRR9856912,PE,H3N2
Flu_2,SRR9856913,PE,H3N2
Ctrl_1,SRR9856914,PE,Mock
Ctrl_2,SRR9856915,PE,Mock

Then, run this (edit DIR_INDEX_ROOT).

DIR_INDEX_ROOT=/dir/to/indeices/created/
VIRTUS_wrapper.py input.csv \
    --genomeDir_human $DIR_INDEX_ROOT/STAR_index_human \
    --genomeDir_virus $DIR_INDEX_ROOT/STAR_index_virus \
    --salmon_index_human $DIR_INDEX_ROOT/salmon_index_human \
    --nthreads=4

You can get this heatmap and summary.csv which contains the ratio of viral reads (hit viral reads/read mapped on the human genome), the stats by Mann-Whitney's U-test, and its false discovery rate.