Installation

System Requirements

  • Python 3.10 or higher

  • 64-bit operating system (Linux, macOS, or Windows)

  • Minimum 8GB RAM (16GB or more recommended)

Basic Installation

Install the latest version from Source:

# Create a new conda environment
conda create -n scExtract python=3.10
conda activate scExtract

# Install scExtract
git clone https://github.com/yxwucq/scExtract
cd scExtract
pip install -e .

Dependencies

Core Dependencies

scExtract depends on the following Python packages:

Package

Version

Description

python

>=3.10 <3.13

Programming language

anndata

>=0.10.7

Data structure for single-cell data

scanpy

>=1.10.1

Single-cell analysis tools

harmonypy

>=0.0.9

Batch effect correction

leidenalg

>=0.10.2

Community detection

louvain

>=0.8.2

Clustering algorithm

mygene

>=3.2.2

Gene annotation

oaklib

>=0.6.3

Ontology analysis

openai

>=1.17.0

OpenAI API integration

pypdf

>=4.2.0

PDF processing

singlecellexperiment

>=0.4.4

Single cell data structure

termcolor

>=2.4.0

Terminal color output

tqdm

>=4.66.2

Progress bars

Optional Dependencies

If you want to use the following features, you will need to install additional packages:

scanorama_prior and cellhint_prior can be installed with git and pip from the following repositories:

git clone https://github.com/yxwucq/scanorama_prior
git clone https://github.com/yxwucq/cellhint_prior

Package

Version

Purpose

anthropic

>=0.25.1

Anthropic API integration

cellhint_prior

Cell type harmonization

scanorama_prior

Batch effect correction

cellhint

>=1.0.0

Cell type harmonization

singler

>=0.4.1

Cell type annotation

scanorama

Batch effect correction

cupy

GPU acceleration