Installation

Local simulation runs the FedGWAS pipeline on one workstation with Flower local simulation mode. It is the recommended first workflow for checking an installation, generating synthetic center data, validating configuration, and comparing federated outputs against a centralized baseline.

Requirements

Use a dedicated Python environment. FedGWAS currently targets:

Dependency	Requirement
Python	3.11 or later
FedGWAS	Installed package or editable repository install
Flower	1.19.x, installed through the FedGWAS dependency set
PLINK	PLINK 1.9 (plink) on PATH or from the repository bundle; PLINK 2 (plink2) for KING-based relatedness screening

PLINK 1.9 is required for federated quality control and logistic-regression association screening. KING-based relatedness screening uses PLINK 2 --make-king-table. The simulation CLI checks for PLINK before starting a run.

Install PLINK

Install PLINK 1.9 so that plink is on PATH, and PLINK 2 as plink2 for relatedness screening. Confirm the installation with:

plink --version
plink2 --version

If PLINK is not on PATH, local simulation projects can still set plink in fedgwas.yaml, but the most portable setup is to expose plink or plink2 in the active environment.

Install From PyPI

For most users, install FedGWAS from PyPI in a clean environment:

python -m venv .venv
source .venv/bin/activate
python -m pip install --upgrade pip
python -m pip install FedGWAS

On Windows PowerShell:

python -m venv .venv
.\.venv\Scripts\Activate.ps1
python -m pip install --upgrade pip
python -m pip install FedGWAS

Confirm that the simulation CLI is available:

fedgwas-sim --help

Verify The Runtime

Run these checks before creating a study:

python --version
python -c "import pipeline; print('FedGWAS import OK')"
python -c "import flwr; print(flwr.__version__)"
fedgwas-sim --help
plink --version

The Flower version should be in the 1.19 line. If it is not, reinstall the supported dependency:

python -m pip install --upgrade "flwr[simulation]>=1.19.0,<1.20"

[Optional] Install from Repository

You can also install FedGWAS directly from the repository, and run simulations with flwr run . local-simulation and local scripts in the repo.

git clone https://github.com/idsla/Fed-GWAS.git
cd Fed-GWAS
python -m pip install -e .

With uv:

git clone https://github.com/idsla/Fed-GWAS.git
cd Fed-GWAS
uv sync --python 3.11

The repository includes bundled PLINK binaries under plink/plink_linux, plink/plink_mac, and plink/plink_win. You can add the appropriate directory to your PATH or set the PLINK path in fedgwas.yaml to use these binaries. And repo contains experiments folder with example configurations and data for local simulation runs.

Difference between PyPI and Repo InstallationL

The repository workflow and the installed-package workflow use the same core pipeline, but their layouts differ:

Workflow	Primary command	Config layout
Installed package	fedgwas-sim ...	User study directory with configs/server.yaml and configs/center_*.yaml
Repository checkout	flwr run . local-simulation	Repository experiments/.../configs/server/config.yaml and center_*/config.yaml

Use one layout consistently within a run.