CasanovoGUI is a cross-platform desktop application for Casanovo — de novo peptide sequencing from MS/MS spectra with a transformer model. Configure your inputs and parameters in a simple form, click Run Casanovo, and watch the output stream live in the console. The GUI wraps Casanovo's command-line sub-commands — de novo sequencing, database search, evaluation, and training — each as its own tab, plus a View tab to explore your results: map peptides to a reference proteome, inspect score distributions and per-residue confidence, and open spectra in PDV for annotated-spectrum visualization.
Download the installer for your platform from the Releases page:
| OS | Download | How to run |
|---|---|---|
| Windows | CasanovoGUI-<version>-windows-x64.zip |
Unzip and run CasanovoGUI.exe |
| macOS (Apple Silicon) | CasanovoGUI-<version>-macos-arm64.dmg |
Open and drag to Applications |
| Linux | CasanovoGUI-<version>-linux-x86_64.deb or …-linux-x86_64.tar.gz |
Install with sudo apt install ./CasanovoGUI-<version>-linux-x86_64.deb, then launch from the desktop menu or run casanovogui; or extract the .tar.gz (no root) and run CasanovoGUI/bin/CasanovoGUI |
No manual tool installation is needed. The installers bundle their own Java
runtime, so you do not have to install Java. And the first time you start an
analysis, if Casanovo isn't already on your machine the GUI offers to install it
for you — it downloads a private Python and Casanovo automatically (no Python or
pip setup required). Just download, launch, and run.
Intel Macs: use the cross-platform
CasanovoGUI-<version>.jarfrom the Releases page instead (it needs a Java 23+ runtime installed).
We have tested CasanovoGUI on the operating systems and computers listed below.
| Configuration | |
|---|---|
| Windows workstation | Windows 11 (build 26200) · Intel Core Ultra 7 265K, 20 cores · 127 GB RAM · NVIDIA RTX 5000 Ada Generation · PyTorch 2.5.1 (CUDA 12.1) |
| Linux server (remote execution) | Ubuntu 22.04.5 LTS · Intel Core i9-10980XE, 18 cores / 36 threads · 125 GB RAM · NVIDIA RTX A4000 16 GB (driver 560.35) |
| macOS laptop (M5) | macOS Tahoe 26.6.2 · MacBook Air (15-inch, M5) · Apple M5 · 16 GB RAM |
| macOS laptop (M4 Pro) | macOS Sequoia 15.7.4 · MacBook Pro (14-inch, 2024) · Apple M4 Pro · 48 GB RAM |
| Windows server | Windows Server 2022 Standard, 21H2 (build 20348) · 2 × Intel Xeon Gold 6354 @ 3.00 GHz · 512 GB RAM |
For a complete guide to using CasanovoGUI — installing it, running each task (de novo sequencing, database search, evaluation, and training), and exploring your results in the View tab — see Getting Started: Graphical User Interface in the Casanovo documentation.
How long a run takes on a GPU or a CPU, and how much memory it needs, is measured for each major Casanovo release in Performance.
Bo Wen, Kai Li, Michael Riffle, Michael J. MacCoss, Wout Bittremieux, William Stafford Noble. "CasanovoGUI: a cross-platform desktop application for deep learning-based de novo peptide sequencing with Casanovo". bioRxiv. 2026.07.11.737889