VirusWarn-Flu

Mutation-Based Early Warning System to Prioritize Concerning Influenza Variants from Sequencing Data

The goal of VirusWarn-Flu is to detect concerning Influenza variants from sequencing data. It does so by parsing Influenza genomes and detecting amino acids mutations in the spike proteins that can be associated with a phenotypic change. The phenotypic changes are annotated according to the knowledge accumulated on previous variants. The tool is based on VirusWarn-SC2, which was invented for SARS-CoV-2.

Documentation

VirusWarn-Flu is part of VirusWarn

Explore »

Getting Started

⚠️ Note: 🔌 Right now, VirusWarn-Flu is tested on Linux and Mac system only 💻

Quick Installation

To run the pipeline, you need to have conda and Nextflow installed and set up. All other dependencies will be installed over conda in the pipeline.

To install conda, use the following bash commands if you are working on Linux:

wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
bash Miniconda3-latest-Linux-x86_64.sh

To install conda, use the following bash commands if you are working on Mac:

wget https://repo.anaconda.com/miniconda/Miniconda3-latest-MacOSX-arm64.sh
bash Miniconda3-latest-MacOSX-arm64.sh

Then, Nextflow an be installed over conda:

conda create -n nextflow -c bioconda nextflow
conda activate nextflow

The VirusWarn-Flu repository can be cloned from Git:

git clone https://github.com/rki-mf1/VirusWarn-Flu.git

Call help

nextflow run main.nf --help

Running VirusWarn-Flu

nextflow run main.nf \
     -profile conda,local \
     --fasta 'test/openflu_h1n1.fasta' \
     --metadata 'test/metadata_h1n1.xlsx'

Running VirusWarn-Flu with splitting (Input from OpenFlu)

nextflow run main.nf \
     -profile conda,local \
     --fasta 'test/openflu_combi.fasta' \
     --metadata 'test/metadata_combi.xlsx' \
     --split 'OpenFlu'

Parameter list

fasta                    REQUIRED! The path to the fasta file with the sequences for VirusWarn-Flu.
                         [ default: '' ]
ref                      If you want to use the recent references from Nextclade, choose ''.
                         H1N1: A/Wisconsin/588/2019 (MW626065)
                         H3N2: A/Darwin/6/2021 (EPI1857216)
                         If you want to use the older references for H1N1 and H3N2, choose 'old'.
                         H1N1: A/California/7/2009 (CY121680)
                         H3N2: A/Wisconsin/67/2005 (CY163680)
                         For Victoria, only B/Brisbane/60/2008 (KX058884) is available.
                         [ default: '' ]
metadata                 The path to a metadate file for the sequences with collection dates.
                         Required to generate a heatmap in the report.
                         [ default: '' ]
subtype                  If the input fasta file only contains sequences of one subtype, 
                         define the subtype to choose the right references and tables.
                         The options are H1N1 and H3N2 for Influenza A,
                         Victoria for Influenza B.
                         [ default: 'h1n1' ]
split                    If the input fasta file contains sequences of more than one subtype, 
                         enable the split parameter to write them into one file per subtype and 
                         ensure the use of the right references and tables.
                         The options are FluPipe, GISAID and OpenFlu.
                         [ default: '' ]
qc                       If set to true, a QC report will be generated from the Nextclade output.
                         [ default: true ]
strict                   Run process with strict alert levels (without orange). Choose 'y'.
                         [ default: 'n' ]
season                   The Influenza season from which the input sequences are.
                         Important for checking on substitutions that are fixed in the population.
                         [ default: '23/24' ]

Data

For further information on the tables that are used for the ranking, please take a look at the subfolders A(H1N1)pdm09, A(H3N2) and B(Victoria) of the folder data depending on the subtype you are interested in and the READMEs that are provided there.

For further instructions for test runs and information on the used data, please take a look at the folder test and the README that is provided there.

An example of the HTML report can be found in the folder example.

How to interprete result.

VirusWarn-Flu output an alert level in four different colours which can be classified into 3 ratings.

Alert color	Description	Impact
Pink	Variants with fixed MOCs and ROIs from the previous season.	MODERATE
Red	Variants with a high number of MOCs and ROIs that can be dangerous.	HIGH
Orange	Variants in the orange level have less MOC and ROI than the ones in the red level and are therefore considered less dangerous but still concerning.	MODERATE
Yellow	Variants that accumulate a high number of ROIs or PMs are sorted in the pink level for further inspection.	MODERATE
Grey	The remaining variants are assigned to the black category.	LOW

Contact

Did you find a bug? 🐛 Suggestion/Feedback/Feature request? 👨‍💻 Please visit GitHub Issues

For business inquiries or professional support requests 🍺 Please feel free to contact us!

Acknowledgments

• Original tool: VirusWarn-SC2 (former VOCAL - Variant Of Concern ALert and prioritization)

  • Original Idea: SC2 Evolution Working group at the Robert Koch Institute in Berlin

  • Funding: Supported by the European Centers for Disease Control [grant number ECDC GRANT/2021/008 ECD.12222].

Citations

An extensive list of references for the tools used by the pipeline can be found in the CITATIONS.md file.