ViroScan is an automated pipeline that eliminate reads not of interest according to a reference (filter-out) and identify viruses present (filter-in).
---
title: Workflow of ViroScan
---
flowchart TD
A([Input reads to analyse]) --> B{Perform a \nfilter out ?};
B -- No --> F{Perform a \nfilter in};
B -- Yes --> C[Align against reference];
D([Indexed reference]) -.-> C;
C --> K([Mapped reads]);
C --> E([Unmapped reads]);
E --> F;
F --> G[Align against viruses];
H[(Database \nof viruses)] -.-> G;
G --> I([Mapped reads\n]);
G --> L([Metrics]);
G --> M([Tool analysis results]);
subgraph identifier["\n\n\nOutput"]
I;
L;
M;
end
Knowing the type of virus present in a sample is essential for certain types of research. In the case of papillomaviruses, a DEIA and LIPA combination is often carried out on the bench to determine whether or not HPVs are present and then to type the sample. However, this approach has certain limitations, notably the financial cost, the time required and the range of HPVs that can be detected. With ViroScan, we are trying to overcome these limitations and make more.
ViroScan create 1 folder log and 3 folders of output :
datawith the reads mapped and unmapped against the reference indexed during the step of filter-out (in orderdata/filter_out/matchedanddata/filter_out/unmatched) and the reads mapped against viruses during the step of filter-in (data/filter_in)analysiswith the output of breseq created during the step of filter-inevidenceswith the tablesfilterin.counts.txt,refs.percents.txtandrefs.coverage.txt
|_ log
|_ output
|_ analysis
|_ data
|_ filter_out
|_ matched
|_ unmatched
|_ filter_in
|_ evidencesHere an exemple of :
- the table filterin.counts.txt
SAMPLE NBR_READS_TO_ALIGN NBR_TOTAL_READS_ALIGNED
sample1 261604 170895
sample2 968370 944474| Field | Description |
|---|---|
| SAMPLE | The name of sample analyse |
| NBR_READS_TO_ALIGN | Number of the reads to analyse into the step filter-in |
| NBR_TOTAL_READS_ALIGNED | Number of the total reads aligned against viruses |
- the table refs.percents.txt
SAMPLE REFERENCE PERCENT_OF_READS_ALIGNED
sample1 FM955841 16.3
sample1 HPU31785 7.3
sample1 PPH5CG 53.6
sample1 X55965 22.0
sample2 HPU31781 99.3| Field | Description |
|---|---|
| SAMPLE | The name of sample analyse |
| REFERENCE | The LOCUS name of the Genbank reference |
| PERCENT_OF_READS_ALIGNED | The percent of reads aligned against the virus |
- the table refs.coverage.txt
SAMPLE REFERENCE AVERAGE_COVERAGE MIN_DEPTH MAX_DEPTH BASES_COVERED_AT_LEAST_TWICE BASES_COVERED_AT_LEAST_TWENTY_TIMES
sample1 FM955841 460.099 0 2614 2097 2095
sample1 HPU31785 204.656 0 1662 2419 2136
sample1 PPH5CG 1497.15 0 7998 5552 4101
sample1 X55965 601.908 0 4321 1542 1539
sample2 HPU31781 7545.88 1 8002 7323 7323| Field | Description |
|---|---|
| SAMPLE | The name of sample analyse |
| REFERENCE | The LOCUS name of the Genbank reference |
| AVERAGE_DEPTH | Average depth |
| MIN_DEPTH | Minimum depth |
| MAX_DEPTH | Maximum depth |
| BASES_COVERED_AT_LEAST_TWICE | Number of covered bases with depth > 2 |
| BASES_COVERED_AT_LEAST_TWENTY_TIMES | Number of covered bases with depth > 20 |
-
Scripts :
-
bash file config.sh into the config directory
-
python script write_output_tables.py into the src directory
-
-
Tools :
- Bowtie2 (http://bowtie-bio.sourceforge.net/bowtie2/index.shtml)
- Python3 (https://www.python.org/downloads/)
- R (https://www.r-project.org/)
- breseq (https://barricklab.org/twiki/pub/Lab/ToolsBacterialGenomeResequencing/documentation/)
- Samtools (http://www.htslib.org/)
NB : Bowtie2 and R are used by breseq
-
Data :
- Viruses references in Genbank format into the directory data
-
Python modules :
- argparse
- json
- re
- sys
First of all, the git project must be cloned into your local directory
git clone https://github.com/srh-bzd/ViroScan.gitThen, move into the folder ViroScan
cd ViroScan/Next, authorise execution of the script ViroScan
chmod +x ./ViroScanTo launch the tool, type the following command in the local directory where the script ViroScan is located
./ViroScan./ViroScan [-h] -f <fastq_directory> -v <virus> -s <se|pe> -o <output_directory> [-x <reference_filter_out>] [-u <1|2>] [-p <threads>] [-b <\"breseq_options\">] [-t <threshold>]| Parameter | Type | Description |
|---|---|---|
| -f | String | Path of the FASTQ file(s) to analyse. In case of paired-end, make sure your files have '_R1*paired', '_R2*paired', '_R1*unpaired' '_R2*unpaired' patterns. The underscore is essential. '.fastq', '.fq' and '.gz' extensions are accepted |
| -v | String | Virus to use to find the closest viral sequence. Available : adenoviridae, adomavirus, anelloviridae, circoviridae, herpesvirales, papillomavirus, parvoviridae, polyomavirus, all. If you want to use your own viral sequences, add a '.gbk' file (Genbank format) into the data directory. Make sure that the name is 'virus.gbk'. You can zip it to limit the space used. In this case, the name of the zip file must be 'virus.gbk.zip' |
| -s | String | Single-end or paired-end data to analyse |
| -o | String | Path of the output directory where results will be written. If the directory doesn't exist, it will be created |
| Parameter | Type | Description |
|---|---|---|
| -x | String | Path of reference indexed and the pattern name of indexes to perform filter out. Make sure the reference was indexed by Bowtie. So, it's means files have to finished by '.bt2' |
| -u | Integer | Number of unpaired files. Default 2 |
| -p | Integer | Number of threads to use. Default 1 |
| -b | String | breseq options. Default "-p -t". Make sure if you want to change default options to input the value with quotation marks. Warning : Do not inform option -j |
| -t | Integer | Threshold value of the minimum alignment percentage to report. Default 5 |
- Ubuntu 18.04.6 LTS - Système d'exploitation
- Visual Studio Code 1.79.2 - Editeur de texte
Sarah BOUZIDI,
Engineer in bioinformatics
Centre National de la Recherche Scientifique (CNRS)
Team Virostyle, Laboratory MIVEGEC, IRD, Montpellier