Home

GEDI

A high-performance, memory-efficient R package for integrating gene expression data from single-cell RNA sequencing experiments. GEDI v2 implements a unified generative model for interpretable latent embedding of multi-sample, multi-condition single-cell data.

devtools::install_github("csglab/gedi2")

Overview

GEDI (Gene Expression Decomposition and Integration) enables:

• Cross-sample integration on par with state-of-the-art methods
• Cluster-free differential expression analysis along the continuum of cell states
• Pathway and regulatory network activity inference in single cells
• Machine learning-based prediction of sample characteristics from single-cell data

Key Features

• Memory-efficient architecture: All data lives in C++ backend (~10GB data with ~1KB R objects)
• High performance: OpenMP parallelization with C++14 optimization
• Sparse matrix support: Efficiently handles sparse single-cell data
• Multiple data modalities: Count matrices (M), paired data (CITE-seq), binary indicators (X), or pre-processed expression (Y)
• Flexible modeling: Dimensionality reduction with batch effect correction

Installation

System Requirements

• R >= 4.0.0
• C++ Compiler with C++14 support
• Eigen >= 3.3.0 (linear algebra library)
• OpenMP (optional, for parallelization)

Installing from GitHub

# Install devtools if not already installed
install.packages("devtools")

# Install gedi from GitHub
devtools::install_github("csglab/gedi2")

System Dependencies

macOS:

brew install eigen

Ubuntu/Debian:

sudo apt-get install libeigen3-dev

CentOS/RHEL:

sudo yum install eigen3-devel

Contributing

Contributions are welcome! Please:

1. Fork the repository
2. Create a feature branch
3. Submit a pull request

Reporting Issues

Report bugs and request features at: https://github.com/Arshammik/gedi/issues

License

This project is licensed under the MIT License - see the LICENSE file for details.