Installation

1. Get the repository

Use a terminal on Linux, macOS, or Windows with WSL. For an existing checkout, change into its directory; otherwise:

git clone https://github.com/MMLabCodes/iPHAsimulatorV2.git
cd iPHAsimulatorV2

2. Create a Python environment

The package requires Python 3.10 or newer; Python 3.11 is used for these guides.

conda create -n ipha python=3.11
conda activate ipha
python -m pip install -e .

This installs the core dependencies declared in pyproject.toml: NumPy, SciPy, NetworkX, pandas, PyYAML and RDKit. Editable installation keeps imports connected to this checkout. No simulation is started by these commands.

3. Add the tools for your workflow

What you want to do	Install
Run notebooks and tests	`python -m pip install -e ".[dev]"`
Use OpenMM, ParmEd and MDTraj	`python -m pip install -e ".[md]"`
Analyse enzyme–PHA contacts	`python -m pip install -e ".[analysis]"`
Launch the Streamlit interface	`python -m pip install -e ".[gui]"`

AmberTools and GROMACS executables are separate from these Python extras. Install the tools your selected tutorial needs, for example through conda-forge:

conda install -c conda-forge ambertools gromacs openbabel

The database builder additionally uses Open Babel’s Python pybel interface. Packmol is needed only for the advanced packing route. GPU platforms and cluster module names depend on your local installation.

4. Check the installation

python -c "import iphasimulator; from iphasimulator.naming import oligomer_name; print(oligomer_name('3HB', 4))"

Expected output: P3HB_4. This checks the package and naming helper; it does not certify every module. For contact analysis, also check:

python -c "import MDAnalysis, matplotlib; from iphasimulator.analysis_contacts import ContactConfig; print('Contact imports OK')"

For 06C, verify tleap is on PATH and import openmm succeeds in the same notebook kernel. For 06A/06B execution, verify gmx --version. Activate the environment before starting Jupyter; a Python import check does not prove that external command-line tools are available.

5. Open a notebook

jupyter lab

Choose a notebook from the catalogue, then select the kernel for the environment you installed. Review its editable file paths and execution switches first. Several notebooks contain machine-specific paths and historical outputs; downloading or viewing them does not validate those inputs for your system.

For building only this documentation, use the separate, lightweight documentation environment. It needs no MD engines.