CUFLynx

A GUI for Circulatory Autogen: run sensitivity analysis, calibration and uncertainty quantification, and manually explore how parameters affect your (CellML) model outputs.

Download the desktop app

OS	Download
Linux (x86-64, glibc 2.35+ / Ubuntu 22.04+)	CUFLynx-linux-x86_64
macOS — Apple silicon (any M-series: M1–M5 and later; macOS 11+)	CUFLynx-macos-arm64
macOS — Intel (macOS 11+)	CUFLynx-macos-x86_64
Windows (x86-64)	CUFLynx-windows-x86_64.exe

Not sure which Mac you have? Apple menu → About This Mac: anything starting "Apple M" is Apple silicon; "Intel…" is Intel. Every M-series chip runs the same arm64 build.

Run it

macOS

Apple silicon:

cd ~/Downloads
chmod +x CUFLynx-macos-arm64
xattr -d com.apple.quarantine CUFLynx-macos-arm64
./CUFLynx-macos-arm64

Intel:

cd ~/Downloads
chmod +x CUFLynx-macos-x86_64
xattr -d com.apple.quarantine CUFLynx-macos-x86_64
./CUFLynx-macos-x86_64

The app isn't notarized yet, so macOS blocks the download until the quarantine flag is cleared — the xattr line does that, or right-click → Open → Open.

Linux

cd ~/Downloads
chmod +x CUFLynx-linux-x86_64
./CUFLynx-linux-x86_64

Windows

Double-click CUFLynx-windows-x86_64.exe. If SmartScreen warns, choose More info → Run anyway.

Windows Defender (or another antivirus) may flag the file as a threat. This is a false positive — a known quirk of the PyInstaller packaging the app uses, not actual malware. If it's quarantined, restore it and allow it in your antivirus, or download again. (The app isn't code-signed yet, which is what would stop this for good.)

These links always point at the newest release. All builds are also listed on the releases page.

The app is self-contained — it bundles Python and everything circulatory_autogen needs, so simulation and analysis (sensitivity / calibration / UQ) run without any Python setup.

On first run, point the app at a circulatory_autogen checkout under Settings → CA dir (clone it with git clone https://github.com/physiomelinks/circulatory_autogen.git). That's the only setup; it's remembered.

Developing circulatory_autogen? You can switch analysis to your own Python under Settings → Python interpreter (pick one with your local CA + its deps installed) instead of the bundled one, so your edits to CA take effect.

Optional: a C compiler (only for the Myokit/CVODE backend)

CUFLynx works out of the box with no compiler. Of the three solver backends, only one needs a C toolchain:

Backend (Settings)	Solver	Needs a C compiler?
python	scipy solve_ivp	no
casadi_python	casadi_integrator	no
cellml_only	CVODE_myokit	yes

Myokit compiles each CellML model to a native extension when it runs, and that toolchain can't be shipped inside the app. If it's missing, CUFLynx shows a warning and you simply pick one of the other two backends. To enable CVODE_myokit:

• Linux — sudo apt install build-essential
• macOS — xcode-select --install
• Windows — Microsoft C++ Build Tools ("Desktop development with C++")

(Sundials/CVODE itself is bundled — you do not need to install it separately.)

Install from source

One-time setup (installs the backend + frontend and builds the UI). Works on Linux, macOS and Windows — needs Python and Node.js installed.

python scripts/install.py

Run

python scripts/run.py

This opens the app at http://localhost:8000.

Calibration / sensitivity / UQ runs use the Python interpreter you pick in the top bar — point it at your circulatory_autogen virtual environment.

Build the desktop app yourself

python scripts/package.py      # -> dist/CUFLynx (or dist/CUFLynx.exe)

Builds the frontend, then freezes the backend + UI + a native window into one executable. PyInstaller can't cross-compile, so build on the OS you're targeting (the release workflow does this on Linux, macOS and Windows runners).

Want a native window without packaging? python apps/desktop/app.py.

---

Developer setup, dev mode (hot reload), API reference and tests: see apps/README.md.