ViralScan

Introduction

ViralScan is a computational bioinformatics framework (Snakemake-based) for detecting viral counts in scRNA-seq samples. It pseudoaligns samples using Kallisto|bustools and performs downstream analyses, such as identifying superexpressors (cells with more than 10 detected viral UMIs) and generating bar plots showing the number of UMIs detected for each virus in a sample.

The tool can use both kb ref and kb count to perform scalable detection of the viral load.

Installation

ViralScan can be installed using pip in a Python environment:

pip install ViralScan

A conda environment is recommended. You can create one as follows:

conda create -n bioenv -c conda-forge -c bioconda snakemake kb-python

ViralScan is command-line based. To view available commands and options, run the following command:

viralscan --help

This will display all available options regarding the software, including an example.

Input Data

Reference Index

ViralScan requires an index created with kb ref (from thhe kb-python package). There are 2 options regarding the index:

• provide your own pre-built index, or;
• let ViralScan create its own index for you.

Reference data to create the index can be found on Zenodo (https://zenodo.org/uploads/16792022).

Samples

ViralScan expects paired-end FASTQ files (potentially gunzipped). Both sample files (forward and backward) must be provided for the analysis. Samples which can be used for testing: SRR20710651, SRR20710645 and SRR10307460

User Guide

There are 2 options regarding the index:

• 1. You already have a reference index built with kb-python
• 2. You don't have a reference index and want ViralScan to create one for you

Please Note: all the output from ViralScan (including the logs and plots) will be created in the output folder defined by the user. To run ViralScan with option 1, run the following:

viralscan -t transcripts.txt -i index.idx -o output/ -s1 sample_1.fastq.gz -s2 sample_2.fastq.gz

If you want ViralScan to create a reference index for you, you have to provide the GTF and FASTA file(s). Running ViralScan to create a reference index and to perform the quantification can be done as follows:

viralscan -o output/ -reference True -s1 sample_1.fastq.gz -s2 sample_2.fastq.gz -fasta fasta.fasta -gtf gtf.gtf

The index will be placed in the output directory, in a subfolder called index.

If you have multiple samples, ViralScan can analyze them with 1 command. Just split the names of the samples with a comma (without a space in-between). The same is when you have multiple GTF and FASTA files. For example:

# If you have multiple samples
viralscan -t transcripts.txt -i index.idx -o output/ -s1 sample1_1.fastq.gz,sample2_1.fastq.gz -s2 sample1_2.fastq.gz,sample2_2.fastq.gz

# If you have multiple gtfs and fasta files 
viralscan -o output/ -reference True -s1 sample_1.fastq.gz -s2 sample_2.fastq.gz -fasta fasta1.fasta,fasta2.fasta -gtf gtf1.gtf,gtf2.gtf

For information about other parameters or possibilities in ViralScan, call the help function:

viralscan --help

License

This project is licensed under the MIT License. See the LICENSE file for details.