Requirements

Python

scCS requires Python ≥ 3.9.

Installation

From PyPI (recommended):

pip install scCS-py

From GitHub (latest development version):

pip install git+https://github.com/mcrewcow/scCS.git

Development install (editable):

git clone https://github.com/mcrewcow/scCS.git
cd scCS
pip install -e ".[dev]"

Core Dependencies

These are installed automatically with scCS:

Package	Version	Purpose
numpy	≥ 1.24	Array math, velocity projections
scipy	≥ 1.10	Statistical tests, sparse matrix support
pandas	≥ 1.5	Results tables, DataFrames
matplotlib	≥ 3.7	All visualizations
seaborn	≥ 0.12	Plot styling
anndata	≥ 0.9	AnnData object support
scanpy	≥ 1.9	Single-cell preprocessing, neighbors
scikit-learn	≥ 1.2	NN entropy computation, GMM fate detection
statsmodels	≥ 0.14	LOWESS smoothing in expression trends

Optional Dependencies

Package	Version	Purpose
scvelo	≥ 0.2.5	RNA velocity computation
gseapy	≥ 1.0	Pathway enrichment (KEGG, GO BP, Reactome)
cellrank	≥ 2.0	CellRank-based fate detection

Install optional dependencies:

pip install scCS-py[velocity]     # scvelo
pip install scCS-py[enrichment]   # gseapy
pip install scCS-py[all]          # everything

Docs Dependencies

To build the documentation locally:

pip install -e ".[docs]"

Includes: sphinx, sphinx-autoapi, furo, nbsphinx, ipykernel.

Tested Environments

OS	Python	Notes
Ubuntu 22.04	3.9, 3.11	Primary development environment
macOS 14	3.10, 3.11	Tested
Windows 11	3.10	Via Anaconda