uMLIP-Interactive: GUI for running simulations with universal machine learning interatomic potentials (uMLIP) - MACE, CHGNet, Nequix, SevenNet, Orb-v3, MatterSim, PET-MAD
Run basic molecular dynamics/static simulations:
- single-point energy
- geometry optimization
- elastic properties
- phonons
- genetic algorithm for finding the energetically most favorible atomic site configurations of substitutes/vacancies
- Basic MD simulations (NVE, NVT, NPT) with generated Python script
- Currently working on: and Nudge elastic band (NEB) calculations.
In an interactive interface for multiple input structure files (POSCAR, CIF, LMP, XYZ with lattice) with a machine learning MACE / CHGNet / Nequix / SevenNet / Orb-v3 / MatterSim / PET-MAD interatomic potentials. It is also possible to generate Python code with the set settings and run the simulations directly in the console (currently works for energies, geometry optimization, elastic properties, and genetic algorithm). For the generated Python code, simply put it into an empty folder and run it. The structures that were uploaded to the app will be automatically created in the POSCAR format before the simulation start. The application also supports selective dynamics in POSCAR files, preserving the set constraints for specific atoms.
- If you like the app, 📖 please cite the following work (arXiv, preprint): Lebeda, M., Drahokoupil, J., Mazáčová, V., & Vlčák, P. (2025). Revealing interstitial energetics in Ti-23Nb-0.7Ta-2Zr gum metal base alloy via universal machine learning interatomic potentials. arXiv preprint.
- Please also 📖 cite the Atomic Simulation Environment (ASE) and 📖 the employed uMLIP (you can see the corresponding publications for each model in the right corner of the app).
- 🎥 Video tutorial for calculations within the interface at YouTube.
- 🎥 Video tutorial for using the genetic algorithm with MACE to obtain candidates for the lowest energy distributions of point defects in crystal structure YouTube.
- 🎥 Video tutorial for using the generated Python script for the calculations at YouTube. Make sure that before running the Python code, you activated the virtual Python environment with the installed necessary Python packages (Step 4 in the compilation manual below).
If you using the app, do not forgot from time to time to update it with 'git pull' in the main folder or download the newest release. I might be adding additional small features from time to time.
If you have any request for improvements, additional features, or you spot a bug, please let me know: [email protected]
Personal recommendation for the maximum number of atoms (for energies, geometry optimization) in a structure file: below 1 000.
If you are using Windows, please compile the app in WSL (if you do not have WSL, see how to install it first e.g. at: https://www.youtube.com/watch?v=zZf4YH4WiZo).
Compile the app: Open your terminal console and write the following commands (the bold text):
(Optional) Install Git:
sudo apt update
sudo apt install git
Currently, certain u-MLIP does not share compatible versions of certain packages (Torch, ...) with each other. Depending on which u-MLIP you want to use, please choose one option:
- Download updates and prerequisites (if you are using WSL, include also 'python3.12-dev'):
sudo apt update
sudo apt install build-essential python3-venv
Download the app code from GitHub (or download it manually without Git on the following link by clicking on 'Code' and 'Download ZIP', then extract the ZIP. With Git, it is automatically extracted):
git clone https://github.com/bracerino/mace-md-gui.git - Navigate to the downloaded project folder:
cd mace-md-gui/ - Create a Python virtual environment to prevent possible conflicts between packages:
python3 -m venv mace_env - Activate the Python virtual environment (before activating, make sure you are inside the mace-md-gui folder):
source mace_env/bin/activate - Install all the necessary Python packages:
pip install -r requirements.txt (with CUDA for calculations on GPU), or (pip install -r requirements-no-cuda.txt (No GPU calculations)) - Run the app (always before running it, make sure to activate its Python virtual environment (Step 4a):
streamlit run app.py - To update the application, write in the main folder:
git pull
- Download updates and prerequisites:
sudo apt update
sudo apt install -y python3-venv
Download the app code from GitHub (or download it manually without Git on the following link by clicking on 'Code' and 'Download ZIP', then extract the ZIP. With Git, it is automatically extracted):
git clone https://github.com/bracerino/mace-md-gui.git - Navigate to the downloaded project folder:
cd mace-md-gui/ - Create a Python virtual environment to prevent possible conflicts between packages:
python3 -m venv mace_env_2 - Activate the Python virtual environment (before activating, make sure you are inside the mace-md-gui folder):
source mace_env_2/bin/activate - Install all the necessary Python packages (for PET-MAD, use requirements-pet-mad.txt):
pip install -r requirements-mattersim.txt (with CUDA for calculations on GPU)) - Run the app (always before running it, make sure to activate its Python virtual environment (Step 4a):
streamlit run app.py - To update the application, write in the main folder:
git pull
- streamlit - 1.46.0
- plotly - 6.1.2
- pymatgen - 2025.6.14
- ase - 3.25.0
- phonopy - 2.40.0
- py3Dmol - 2.5.0
- psutil - 7.0.0
- gputil - 1.4.0
- torch - 2.7.1
- torchvision - 0.22.1
- torchaudio - 2.7.1
- mace-torch - 0.3.13
- cuequivariance - 0.5.1
- cuequivariance-torch - 0.5.1
- cuequivariance-ops-torch-cu12 - 0.5.1
