Peer-reviewed · GigaScience 2022 Nextflow DSL2 Docker & Singularity v2.0 rework

Identify, characterize & understand your phages in one run.

What the Phage (WtP) v2.0 combines 16 established phage-identification tools with quality control, annotation, taxonomy, prophage detection, host prediction and lifecycle prediction - then delivers everything in a single interactive report.

16+ phage-tools
7 analysis modules
1 command
1 interactive report
0
Phage-ID tools
0
Analysis modules
0
Taxonomy tools
0
Command to run it all
Pipeline

From contigs to characterization - automatically

Every module is an independent Nextflow workflow. Run them together with --end_to_end, or invoke only the steps you need.

FASTA input

.fa / .fasta / .fna / .gz contigs

Input validation

Header sanitizing + length filter (default 1500 bp)

1

Identification

16 tools, normalized viral scores

2

Quality control

CheckV completeness & contamination

3

Annotation

Prodigal · Pharokka · GeNomad · PhaBox2

4

Taxonomy

Sourmash · GeNomad · PhaGCN · tax my phage

5

Prophage detection

GeNomad · PhaBox2 · VirSorter2 · Phigaro

6

Host prediction

iPHoP · CHERRY

7

Lifecycle

BACPHLIP · PhaTYP · consensus

Interactive report

One self-contained HTML report with tabs, genome viewer, per-sample JSON and more.

All 7 modules run in parallel on the validated input.
Run everything with --end_to_end, or pick modules: --identify --annotation --taxonomy --prophage --host --lifecycle
Tools

16+ established tools, orchestrated for you

WtP wraps, runs and filters a panel of established phage tools so you get clean, comparable results.

Quality control

Viral genome completeness & contamination

Host prediction new

Which bacterium does the phage infect?

Phage-therapy safety in development

Screening for genes of concern before therapy use
Click any tool to visit its repository Raw tool outputs archived for transparency Per-contig x per-tool agreement matrix included
Features

Built for wet labs, bioinformaticians & phage therapy

Reproducible, containerized and scaleable - from a laptop to a compute cluster or the cloud.

One command, full stack

--end_to_end chains every module automatically: identify, QC, annotate, classify, detect prophages, predict host and lifecycle.

Modular by design

Independent workflow flags let you run only what you need:

  • --identify · --annotation
  • --taxonomy · --prophage
  • --host · --lifecycle

Databases, automatic

Reference databases auto-download on first run and are available for subsequent runs - they don't need to be downloaded every time.

Interactive HTML report

Overview, quality, identification, annotation genome viewer (linear & circular), taxonomy, prophage and lifecycle - plus per-sample JSON for downstream use.

Scales: laptop → HPC → cloud

Profiles for local, slurm, lsf, ebi and Google Batch, with docker or singularity engines.

Reproducible

Containerized tools, pinned Nextflow and CI-tested stub profiles make every run reproducible and fast to test.

Installation

Up and running in a few commands

WtP requires Java, Nextflow and a container engine (Docker or Singularity). This is the Ubuntu quick-start path.

Databases · end-to-end

≈ 50 GB

auto-downloaded on first run

Optional add-ons

+≈ 850 GB iPHoP

+≈ 32 GB PhageScope

only with --iphop / --phage_scope_tax

Test profile smalltest

≈ 1.4 GB

minimal databases, runs in ~1 h

Database sizes per tool

StepDatabaseSize
IdentificationVirSorter13 GB
IdentificationVirSorter211 GB
IdentificationVIBRANT3.3 GB
IdentificationSourmash (identify)~260 MB
IdentificationMetaPhinder~300 MB
IdentificationPPR-Meta~40 MB
QC · completenessCheckV5.4 GB
AnnotationpVog + VOG table~1.6 GB
TaxonomySourmash GTDB14 GB
TaxonomygeNomad800 MB
TaxonomyNCBI Taxonomy~320 MB
Prophagereuses VirSorter2 / geNomad / Phigaro—
HostPhaBox2—
Host --iphopiPHoP~850 GB
Taxonomy --phage_scope_taxPhageScope (~800k sequences)~32 GB

Plus container images (several GB, depending on Docker/Singularity). geNomad peaks at ~20 GB RAM.

Run times · real end-to-end run

ModuleTime
Total (8 assemblies, 20 cores)≈ 19 h
Identification (16 tools)~11.5 h
Prophage detection~4.5 h
Taxonomy~1 h
Annotation~1 h
Lifecycle~40 min
CheckV completeness~15 min
Host prediction~10 min

Measured on 8 clinical assemblies running --end_to_end with 20 cores. Slowest single tools per assembly: VirSorter2 ~29 min, VirSorter ~21 min, VIBRANT ~6 min, Pharokka ~5 min. Times scale with the number and size of your input assemblies.

Install Java and Nextflow
# install java and nextflow
sudo apt-get update
sudo apt install -y default-jre
curl -s https://get.nextflow.io | bash
sudo mv nextflow /usr/bin/
Install Docker Engine
# Add Docker's official GPG key:
sudo apt update
sudo apt install ca-certificates curl
sudo install -m 0755 -d /etc/apt/keyrings
sudo curl -fsSL https://download.docker.com/linux/ubuntu/gpg -o /etc/apt/keyrings/docker.asc
sudo chmod a+r /etc/apt/keyrings/docker.asc

# Add the repository to Apt sources:
sudo tee /etc/apt/sources.list.d/docker.sources <<EOF
Types: deb
URIs: https://download.docker.com/linux/ubuntu
Suites: $(. /etc/os-release && echo "${UBUNTU_CODENAME:-$VERSION_CODENAME}")
Components: stable
Architectures: $(dpkg --print-architecture)
Signed-By: /etc/apt/keyrings/docker.asc
EOF

sudo apt update
sudo apt install docker-ce docker-ce-cli containerd.io docker-buildx-plugin docker-compose-plugin

sudo usermod -a -G docker $USER
Restart your computer so the Docker group takes effect, then run a quick testrun using the stable release (-r v2.0.0):
Test run
# -r v2.0.0 = stable release version (pinned for reproducibility)
nextflow run replikation/What_the_Phage -r v2.0.0 --cores 8 -profile smalltest,local,docker
How to run

Three commands cover the spectrum

Requires Nextflow and a container engine. Databases auto-download on first run (≈ 50 GB for the default end-to-end run). Add -resume to cache completed steps.

End-to-end on a laptop
# full characterization: identify + QC + annotate + taxonomy + prophage + host + lifecycle
# phage.nf = newest changes, run inside the cloned What_the_Phage directory (may be unstable)
# or use the stable release from anywhere: nextflow run replikation/What_the_Phage -r v2.0.0 ...
nextflow run phage.nf \
    --fasta 'contigs/*.fasta' \
    --end_to_end \
    --cores 16 \
    --output results \
    -profile local,docker
HPC cluster (SLURM / Singularity)
# phage.nf = newest changes, run inside the cloned What_the_Phage directory (may be unstable)
# or use the stable release from anywhere: nextflow run replikation/What_the_Phage -r v2.0.0 ...
nextflow run phage.nf \
    --fasta 'contigs/*.fasta' \
    --end_to_end --cores 20 \
    --output results \
    -profile slurm,singularity \
    --cachedir /images/singularity_images \
    --databases /databases/WtP_databases/
Modular - run only what you need
# e.g. identification + prophage detection only
# phage.nf = newest changes, run inside the cloned What_the_Phage directory (may be unstable)
# or use the stable release from anywhere: nextflow run replikation/What_the_Phage -r v2.0.0 ...
nextflow run phage.nf \
    --fasta 'contigs/*.fasta' \
    --identify --prophage \
    --output results \
    -profile lsf,singularity

Executors

local · slurm · lsf · ebi · Google Batch (ukj_cloud) · test · stub

Engines

docker or singularity - always pick one: -profile EXECUTER,ENGINE

Input

Assembled contigs, .fa / .fasta / .fna / .gz. Multiple files are auto-split into per-sample results.

Tool options

Everything phage.nf --help knows

The full command-line reference for WtP v2.0. Run nextflow run phage.nf --help on any machine to get this list.

phage.nf --help
# standard execution
nextflow run phage.nf --fasta '*/*.fasta' --cores 20 \
    --end_to_end --output results -profile local,docker

# modular execution - run only what you need
nextflow run phage.nf --fasta '*/*.fasta' --cores 20 \
    --identify --prophage --output results -profile lsf,singularity

# input
--fasta    assembly file(s), e.g. '*.fasta'
--setup   skip analysis, download databases + containers

# workflow control
--end_to_end  identify + annotate + taxonomy + prophage + host + lifecycle
--identify    run phage identification
--annotation  run annotation
--taxonomy    run taxonomy
--prophage    run prophage detection
--host        run host prediction
--lifecycle   run lifecycle prediction

# profiles & engines:  -profile <Executor>,<Engine>
executors   local | slurm | lsf | ebi | ukj_cloud
engines     docker | singularity
smalltest   test environment (~1h):  -profile smalltest,local,docker

Input

--fastaassembly file(s) [required, e.g. '*.fasta']
--setupskip analysis, download databases + containers

Workflow control

--end_to_endidentification + annotation + taxonomy + prophage + host + lifecycle
--identifyrun phage identification
--annotationrun annotation
--taxonomyrun taxonomy
--prophagerun prophage detection
--hostrun host prediction
--lifecyclerun lifecycle prediction

General options

--filterminimum contig size in bp to keep [1500]
--corescores per process
--max_corestotal cores used on the machine
--memmemory in GB [12]
--outputname of the results folder [results]
--all_toolsactivate all phage prediction tools
--iphopenable iPHoP host prediction

Tool controls (deactivate)

--dvdeactivate DeepVirFinder
--mpdeactivate MetaPhinder
--ppdeactivate PPR-Meta
--smdeactivate Sourmash
--vbdeactivate VIBRANT
--vfdeactivate VirFinder
--vndeactivate VirNet
--vsdeactivate VirSorter
--phdeactivate Phigaro
--vs2deactivate VirSorter2
--skdeactivate Seeker
--pb2deactivate PhaBox2 - PhaMer

Visualization & databases

--pharokkaactivate Pharokka plot
--plot_completenessplot contigs with CheckV completeness above this [75.00]
--phage_scope_taxswitch to curated PhageScope DB [~800k sequences]
--annotation_dbpath to custom annotation HMM database

System paths

--databasesdatabase download location [nextflow-autodownload-databases]
--workdirtemporary work directory
--cachedirsingularity cache directory [singularity_images]
Live demo

See a real result explorer

Three demo datasets are bundled with this site. Browse tabs, switch samples, explore genes.

Interactive result explorer - 3 real samples, 7 tabs (overview, quality, identification, taxonomy, annotation, prophage, lifecycle)
Publication

Peer-reviewed & open

Giga
Science

What the Phage: a scalable workflow for the identification and analysis of phage sequences

M. Marquet, M. Hölzer, M. W. Pletz, A. Viehweger, O. Makarewicz, R. Ehricht, C. Brandt

Peer-reviewed in GigaScience (2022) · open source under GPL-3.0 · actively developed with release candidates

doi:10.1093/gigascience/giac110

Ready to find out what your phages are?

Clone the repo, run one command, and open the interactive report. Reports, issues and feature requests are welcome on GitHub.