https://github.com/aziele/phirbo

Download taxdb.btd, taxdb.bti, taxdb.tar.gz from https://ftp.ncbi.nlm.nih.gov/blast/db/taxdb.tar.gz
Prepare rank-biased overlap blast results via phirbo_preprocessing.txt
Run phribo by providing two input directories (i.e., for phages [phage_virusblast/] and bacteria [phage_hostblast/]) containing ranked lists from blast output, and an output file name (phage_phirbo/predictions.csv)

python phirbo/phirbo.py phage_virusblast/ phage_hostsblast/ phage_phirbo/predictions.csv

https://github.com/refresh-bio/PHIST

Due to dependencies, PHIST will be run using a Singularity container image. To create the image, run the following commands:

# Build the image: apptainer build phist.sol.sif phist_image.def # Create an interactive session using the PHIST image: SIMG=phist.sol.sif interactive # Activate the base environment: cd /opt/miniconda/bin . activate 

PHIST takes as input two directories containing FASTA files (gzipped or not) with genomic sequences of viruses (phage_genome_dir) and candidate hosts (host_dir)
python PHIST/phist.py phage_genome_dir host_dir phage_PHIST_outdir
Results will be in the output directory (phage_PHIST_outdir) labeled as predictions.csv

https://github.com/congyulu-bioinfo/PHP

First, calculate the K-mer frequency of the host
python3 PHP/countKmer.py -f phage_host_genomes -d phage_host_PHPkmer -n phage_PHPHostKmer -c -1 where:

• -f phage_host_genomes: fasta file of prokaryotic genome sequence (one fasta per file)
• -d phage_host_PHPkmer: path of the prokaryotic k-mer file
• -n phage_PHPHostKmer: name of the prokaryotic k-mer file
• -c -1: the number of cores used; -1 represents the use of ALL cores

Then, predict the infection relationship between the virus and the host
python3 PHP/PHP.py -v phage_genomes -o phage_PHPout -d phage_PHPkmer -n phage_PHPHostKmer

• -v phage_genomes: fasta file of query virus sequences (one virus genome per file)
• -o phage_PHPout: path of temp and result files
• -d phage_PHPkmer: path of the prokaryotic k-mer file
• -n phage_PHPHostKmer: name of the prokaryotic k-mer file

https://github.com/jessieren/VirHostMatcher

To run VHM create a folder containing virus fasta files and a folder containing host fasta files - no subfolders
python VirHostMatcher/vhm.py -v phage_genomes/ -b phage_host_genomes/ -o Phage_vhm_output/ -d 1

• -v phage_genomes/: path to virus folder
• -b phage_host_genomes/: path to host folder
• -o Phage_vhm_output/: path to output
• -d 1: only calculate the d2* dissimilarity (1 for yes, 0 for no; default: 0)

https://github.com/soedinglab/WIsH

Null directory is created from a diverse range of Alteromonas, Cellulophage, Cyanophage, Lactobacillus, Mycobacterium, Oenococcus, Pelagibacter, Prochlorococcus, Rhizobium, Synechococcus, and Thermus phage genomes that are known not to infect the bacterial model listed in null.txt
Create host model directory
WIsH/WIsH -c build -g phage_hosts_genomes/ -m modelDir

• -c build: tell WIsH to create a model
• -g phage_hosts_genomes/: where bacterial genomes in FASTA format are stored
• -m modelDir: where model for each bacterial genome will be stored

Run predict on null to get llikelihood.matrix in outputNullModelResultDir to feed into computeNullParameters.R

WIsH/WIsH -c predict -g null/ -m modelDir -r outputNullModelResultDir -b 1000 Rscript ../WIsH/computeNullParameters.R WIsH/WIsH -c predict -g phage_genomes/ -m modelDir -r outputResultDir -b 1000 -n outputNullModelResultDir/nullParameters.tsv 

• -c predict: tell WIsH to run a prediction
• -g null/: where the null phage genomes in FASTA format are stored
• -g phage_genomes/: where query phage genomes in FASTA format are stored
• -m modelDir: where model for each bacterial genome will be stored
• -r outputResultDir: where results will be stored
• -b 1000: k option to output the k best prediction by likelihood
• -n outputNullModelResultDir/nullParameters.tsv: location of null model; to be used when p-values are desired

Exploratory tool analyses were done on the Open Science Pool from the Open Science Grid (https://osg-htc.org/services/open_science_pool.html) and ASU's Sol supercomputer

All exploratory tools require a docker image to be created before running the tools

apptainer build toolname.sol.sif toolname.def apptainer shell toolname.sol.sif cd /opt/miniconda/bin . activate 

The cd /opt/miniconda/bin and . activate commands are important in activating the underlying conda environment within the container.

https://github.com/KennthShang/CHERRY

To run CHERRY, do the following:

• Clone the CHERRY repository to a directory: git clone https://github.com/KennthShang/CHERRY.git
• Run a shell on the built containder using ASU's Sol supercomputer (which uses Slurm): SIMG=cherry.sol.sif interactive
• Activate the environment: cd /opt/miniconda/bin and . activate
• Change directory to where you cloned the CHERRY repository: cd CHERRY/
• Prepare the database:

cd dataset bzip2 -d protein.fasta.bz2 bzip2 -d nucl.fasta.bz2 cd ../prokaryote gunzip * cd .. 

Predict hosts for viruses using a fasta file containing the viral sequences as the input

python run_Speed_up.py --contigs all.fasta --model pretrain --topk 1000

• --contigs: input fasta file (all.fasta is a single FASTA with multiple sequences)
• --model: predicting host with pretrained (or retrained) parameters (default: pretrained)
• --topk: host prediction with topk score (default: 1)

https://github.com/KennthShang/HostG

To run HostG, do the following:

• Clone the HostG repository to a directory: git clone https://github.com/KennthShang/HostG.git
• Run a shell on the built containder using ASU's Sol supercomputer (which uses Slurm): SIMG=hostg.sol.sif interactive
• Activate the environment: cd /opt/miniconda/bin and . activate
• Change directory to where you cloned the HostG repository: cd HostG/
• Prepare the database (similar to CHERRY):

cd HostG/dataset bzip2 -d protein.fasta.bz2 bzip2 -d nucl.fasta.bz2 

Predict hosts for viruses using a fasta file containing the viral sequences as the input

python run_Speed_up.py --contigs all.fasta --t 0

• --contigs: path to the contigs file (all.fasta is a single FASTA with multiple sequences)
• --t: threshold value; predictions will only be outputted that have confidence (SoftMax) values above t (default: 0)

https://github.com/felipehcoutinho/RaFAH

To run RaFAH, do the following:

• Clone the RaFAH repository to a directory: git clone https://github.com/felipehcoutinho/RaFAH.git

• Run a shell on the built containder using ASU's Sol supercomputer (which uses Slurm): SIMG=rafah.sol.sif interactive

• Activate the environment: cd /opt/miniconda/bin and . activate

• Change directory to where you cloned the RaFAH repository: cd RaFAH/

• Run perl RaFAH.pl --fetch to download all necessary input files

• Edit the perl script to include the path to the pretrained model files:

  Line 28: Full path to HP_Ranger_Model_3_Valid_Cols.txt Line 29: Full path to HP_Ranger_Model_3_Filtered_0.9_Valids.hmm Line 30: Full path to RaFAH_Predict_Host.R Line 31: Full path to RaFAH_train_Model.R Line 32: Full path to MMSeqs_Clusters_Ranger_Model_1+2+3_Clean.RData

Perform host predictions using a set of pre-computed model files

perl RaFAH.pl --predict --genomes_dir phage_genomes/ --extension .fasta

• --genomes_dir: directory containing DNA sequences in FASTA format
• --extension: extension of the FASTA files to be analyzed (e.g., fasta, fna, fa; default: fasta)

https://github.com/LaboratorioBioinformatica/vHULK

To run vHULK, do the following:

• Clone the vHULK repository to a directory: git clone https://github.com/LaboratorioBioinformatica/vHULK
• Run a shell on the built containder using ASU's Sol supercomputer (which uses Slurm): SIMG=vhulk.sol.sif interactive
• Activate the environment: cd /opt/miniconda/bin and . activate
• Change directory to where you cloned the RaFAH repository: cd vHULK/

Run vHULK on target genera

python vHULK.py -i input/ -o output/ --all

• -i: input directory; path to a folder containing metagenomic bins in .fa or .fasta format
• -o: output directory; location to store results in -- will be created if absent
• --all: write predictions for all input bins/genomes, even if they were skipped (i.e., size filtered or hmmscan failed)

https://github.com/WeiliWw/VirHostMatcher-Net

To run VirHostMatcher-Net, do the following:

• Run a shell on the built containder using ASU's Sol supercomputer (which uses Slurm): SIMG=vhmn.sol.sif interactive
• Activate the environment: cd /opt/miniconda/bin and . activate
  Note: VirHostMatcher-Net is already built into the container image

Run VirHostMatcher-Net
python /opt/VirHostMatcher-Net/VirHostMatcher-Net.py -q input -o output -i tmp -n 10 -t 10

• -q: directory containing query virus genomes with .fasta or .fa extension
• -o: output directory
• -i: directory storing intermediate results
• -n: topN; number of top predictions written to the output files (all predictions will be output if there is a tie in score)
• -t: number of threads to use

https://github.com/biocom-uib/vpf-tools

To run VPF-Class, do the following:

• Clone the VPF-Class repository to a directory: git clone https://github.com/biocom-uib/vpf-tools
• Run a shell on the built containder using ASU's Sol supercomputer (which uses Slurm): SIMG=vpfclass.sol.sif interactive
• Activate the environment: cd /opt/miniconda/bin and . activate
• Change directory to where you cloned the RaFAH repository: cd vpf-tools/
• Build the tool: stack build
• Download the supplementary material: wget https://bioinfo.uib.es/~recerca/VPF-Class/vpf-class-data.tar.gz
• Download HMMER

wget http://eddylab.org/software/hmmer/hmmer-3.4.tar.gz tar -xvzf hmmer-3.4.tar.gz cd hmmer-3.4 ./configure make 

Run VPF-Class

stack exec -- vpf-class \ --input-seqs all.fasta \ --output-dir out \ --data-index vpf-class-data/index.yaml 

• --data-index: file that specifies classification levels
• -i: input file (all.fasta is a single FASTA with multiple sequences)
• -o: output directory