Skip to content

Repository files navigation

🦠 Phang

End-to-end bacteriophage genome characterisation — from FASTA to a phage-therapy verdict, in one click.

Latest release Platform License: MIT Downloads

Phang takes one or more bacteriophage genome FASTA files and runs a 12-step pipeline across nine specialised tools to annotate genes, predict lifestyle and host, place the phage in viral taxonomy, screen for safety red-flags, and judge phage-therapy suitability — then packages everything into an interactive HTML report card and a submission-ready NCBI BankIt file.

No terminal, no conda setup: download the macOS installer, double-click, and run.


✨ Highlights

  • One-click desktop app for Apple Silicon Macs — no command line needed.
  • Phage-therapy verdict — flags whether a phage is a candidate (strictly lytic, and free of antimicrobial-resistance / virulence / lysogeny genes), and shows the evidence behind every call.
  • Interactive report card — genome map, gene annotations, lifestyle, predicted host, viral taxonomy, defense systems, depolymerases, and receptor-binding proteins.
  • Batch dashboard — analyse many phages at once and compare them at a glance.
  • NCBI BankIt package — submission-ready FASTA + feature table.
  • Nine tools, one workflow — Pharokka, Phold, Phynteny, PhaStyle, PhaBOX2/CHERRY, vConTACT3, DefenseFinder, PhageRBPdetect, DepoScope.

💻 Requirements

  • Apple Silicon Mac (M1 or later) — runs natively, no Rosetta
  • macOS 11 or later
  • ~35 GB free disk space (a one-time ~30 GB database download on first launch)
  • Internet connection for the first-run setup

⬇️ Installation

  1. Download Phang-<version>-MacOSX-arm64.pkg from the latest release and double-click it.
  2. This beta is unsigned, so macOS Gatekeeper blocks it the first time:
    • macOS 15 (Sequoia) or later — on the "cannot be opened…" warning click Done, then open System Settings → Privacy & Security, scroll to Security, and click Open Anyway.
    • macOS 11–14 — right-click (Control-click) the file in Finder → OpenOpen.
  3. Click through the installer (Continue → Install; it asks for your Mac password). Phang installs to /Applications.
  4. Open Phang from Applications or Launchpad. The app is also unsigned, so apply the same Gatekeeper step once the first time you open it.

🚀 First launch

On first launch, Phang downloads its analysis tools and reference databases (~30 GB) into ~/.phang. A progress window shows the status — keep it open. You can quit to pause; finished downloads are kept and it resumes next time.

🧬 Usage

  1. Open Phang.
  2. Click Choose folder… and select a FASTA file or a folder of FASTA files.
  3. Choose an output folder.
  4. Click Run Pipeline.

Accepted inputs: a single FASTA file, a folder of FASTA files, or a multi-record FASTA. Each phage gets its own report_card.html, and a batch_dashboard.html links them all together.

📑 What you get

<output>/
├── <phage_name>/
│   └── report_card.html      ← interactive per-phage report
├── batch_dashboard.html      ← overview of every phage
├── ncbi/
│   ├── bankit_multi.fasta    ← NCBI BankIt submission
│   └── bankit_features.tbl
└── _logs/                    ← per-step logs

⚙️ The pipeline

# Tool What it does Typical time
1 Pharokka Gene calling + functional annotation (PHROG, CARD, VFDB) ~2 min
2 Phold Structure-based annotation (ProstT5 + Foldseek) ~36 s
3 Phynteny Synteny-based annotation of hypothetical genes <5 s
4 PhaStyle Lifestyle prediction (virulent / temperate) ~18 s
5 PhaBOX2 / CHERRY Host prediction ~34 s
6 vConTACT3 Viral taxonomy (genus / family) ~4 min
7 DefenseFinder + PhageRBPdetect Anti-defense systems + receptor-binding proteins ~10 s
8 DepoScope Depolymerase detection ~21 s
9–11 Genome viz · NCBI · Report Genome map, BankIt package, report card ~11 s
Total (single genome) ~8–9 min

💊 How the phage-therapy verdict works

A phage is labelled a candidate when it is strictly lytic and shows no antimicrobial-resistance genes, no virulence factors, and no integrase / recombinase (lysogeny markers). Antimicrobial-resistance and virulence are judged by Pharokka sequence homology against CARD and VFDB (≥80% identity); low-identity structural look-alikes from Phold are shown as informational notes and do not count against a phage. Every call on the report card has a hover tooltip explaining how it was derived.

🙏 Acknowledgements

Phang orchestrates these excellent open-source tools — please cite them if you use Phang in your work: Pharokka, Phold, Phynteny, PhaStyle (ProkBERT), PhaBOX2 / CHERRY, vConTACT3, DefenseFinder, PhageRBPdetect, and DepoScope.

📄 License

MIT © seanpang49


Phang is research software provided as an unsigned beta. Results are predictions and should be confirmed experimentally before any clinical or therapeutic decision.

Releases

Packages

Contributors

Languages