VirHostMatcher-Net is a network-based computational tool for predicting virus-host interactions. Current version predicts hosts of given viruses from a database of 62,493 Bacteria and Archaea candidates. VirHostMatcher-Net has two modes: predicting for complete genomes and predicting for short viral contigs.
Important: We corrected a critical bug in the software: the coefficients for SV+ and SV- (see Equation 6 in the paper) were mistakenly swapped when the software was published. Note this does not impact the results in the paper: we computed the feature values separately without the software (because the computation/simulation used in the study was too time-consuming for a single run by the software, especially for the blast feature that we evaluated in the study), and directly applied the coefficients to calculate the final score.
Wang et al. "A network-based integrated framework for predicting virus–prokaryote interactions" NAR Genomics and Bioinformatics, Volume 2, Issue 2, June 2020, lqaa044, https://doi.org/10.1093/nargab/lqaa044.
VirHostMatcher-Net requires Python 3.4+ together with the following packages and BLAST.
We recommend to use Miniconda to install all dependencies. After installing Miniconda, simply run
conda install numpy pandas Biopython
conda install -c bioconda blast # skip this line if you've already installed BLAST
Alternatively, users may want to install Python3.X and dependencies manually.
In addition to dependencies, VirHostMatcher-Net requires to build local modules and download database.
git clone https://github.com/WeiliWw/VirHostMatcher-Net.git
cd VirHostMatcher-Net
CC=g++ python setup.py install --install-platlib=./src/
## Optional
# Include the VirHostmatcher-Net directory in your $PATH
# The main script is executable
export PATH=/path/to/VirHostMatcher-Net/:$PATH
VirHostMatcher.py -h
git clone https://github.com/WeiliWw/VirHostMatcher-Net.git
cd VirHostMatcher-Net
MACOSX_DEPLOYMENT_TARGET=10.9 CC=g++ python setup.py install --install-platlib=./src/
The prediction model of VirHostMatcher-Net depends on a large amount of data: BLAST index files of all bacteria and their CRISPRs, WIsH models(short viral contig mode) and hash files for calculating s*2, etc.
There are two packs of files for downloading, based on your use case: if you intend to use the software on complete genomes or long contigs, you can download "Pack 1" below which is much smaller than "Pack 2". The data can be directly downloaded from Google Drive, or you can use curl in the commandline to download the files from Google Drive (see below, credit to this genius post).
At the directory of VirHostMatcher-Net, run
# download
fileid="185U3ZLYe1uNmB5oCaIlb0IQNURnlyXcN"
filename="data_VirHostMatcher-Net_complete_genome_mode_alone.tar.gz"
curl -c ./cookie -s -L "https://drive.google.com/uc?export=download&id=${fileid}" > /dev/null
curl -Lb ./cookie "https://drive.google.com/uc?export=download&confirm=`awk '/download/ {print $NF}' ./cookie`&id=${fileid}" -o ${filename}
# extract
tar xf data_VirHostMatcher-Net_complete_genome_mode_alone.tar.gz
Note: The extracted
datafolder takes up 125G of disk space.
At the directory of VirHostMatcher-Net, run
# download
fileid="1ZWTn_WIkSbtr6guyAPNnmmYai-qmh93P"
filename="data_VirHostMatcher-Net_both_modes.tar.gz"
curl -c ./cookie -s -L "https://drive.google.com/uc?export=download&id=${fileid}" > /dev/null
curl -Lb ./cookie "https://drive.google.com/uc?export=download&confirm=`awk '/download/ {print $NF}' ./cookie`&id=${fileid}" -o ${filename}
# extract
tar xf data_VirHostMatcher-Net_both_modes.tar.gz
VirHostMatcher-Net accepts files in FASTA format.
python VirHostMatcher-Net.py [-h] -q QUERY_VIRUS_DIR [-t NUM_THREADS] [--short-contig] -o
OUTPUT_DIR [-n topN] [-i INTERMEDIATE_DIR] [-l GENOME_LIST]
-h, --help show this help message and exit
-q QUERY_VIRUS_DIR Directory containing query virus genomes with .fasta or
.fa suffix
-t NUM_THREADS Number of threads to use. Default = 1
--short-contig Predict hosts for short viral contigs. WIsH model files
are required in this mode
-o OUTPUT_DIR Output directory
-n topN Number of top predictions written to the output files.
All predictions will be output if there is a tie in
score. Default = 1
-i INTERMEDIATE_DIR Directory storing intermediate result. Default =
./intermediate_res
-l GENOME_LIST Location of the file containing host NCBI genome names of
interest
mkdir output
python VirHostMatcher-Net.py -q ./test/VGs -o output -i tmp -n 3 -t 8
mkdir output2
python VirHostMatcher-Net.py -q ./test/mVCs --short-contig -o output2 -n 3 -t 8 -l genome_list/marine_host_list.txt
In both modes, VirHostMatcher-Net outputs a prediction file for each query virus to the specified directory. A prediction file is in .csv format where each row represents one candidate host with detailed taxanomic information, a prediction score, values (_val) and percentiles (_pct) of each feature. The feature percentile of a virus-host pair is defined as the percentile of this feature score among all scores between that virus and all the candidate hosts. A very high percentile (i.e. >95%) suggests significance of the feature on contributing to the prediction. In the output, the percentile of SV-, with a negative coefficient, is reversed to keep consistent with other feature percentiles. Tables of feature values are stored in a subdirectory feature_values under the output directory.
Users can use a subset of candidate hosts for prediction by the option -l to specify a list of NCBI genome names. Two example lists of 4034 marine hosts and 9097 human-associated hosts can be found in the directory genome_list.
Training and validation data are shared in Google Drive.
Please open a Github issue or contact Weili Wang weiliw@usc.edu
Copyright (C) 2019 University of Southern California, Weili Wang and Fengzhu Sun
This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this program. If not, see http://www.gnu.org/licenses/.