🧬 scGraphVerse
✨ Comprehensive Gene Regulatory Network Analysis for Single-Cell Data ✨
🎯 Overview
scGraphVerse is a comprehensive R package for inferring, evaluating, and visualizing gene regulatory networks (GRNs) from single-cell RNA sequencing data. It provides an integrated framework with multiple inference algorithms, consensus construction, and rich visualizations optimized for single-cell expression analysis.
⚡ Key Features
| Feature | Description |
|---|---|
| 🔬 Multiple Inference Methods | GENIE3, GRNBoost2, ZILGM, JRF, PCzinb |
| 🤝 Consensus Networks | Voting, union, INet integration |
| 📊 Comprehensive Evaluation | ROC curves, AUC, F1-score, community analysis |
| 🎨 Visualizations | Interactive networks, performance plots |
| 🔧 Flexible Integration | SingleCellExperiment, Seurat, matrix objects |
🚀 Installation
💻 Development Version
# Install development version
if (!require("devtools")) install.packages("devtools")
devtools::install_github("ngsFC/scGraphVerse")🔬 Inference Algorithms
| Method | Description |
|---|---|
| GENIE3 🌳 | Tree-based ensemble learning |
| GRNBoost2 🚀 | Gradient boosting with Dask |
| ZILGM 🎯 | Zero-inflated Gaussian graphical models |
| JRF 🌲 | Joint Random Forests |
| PCzinb 🔗 | Partial correlation with ZINB |
🎪 Quick Start Demo
library(scGraphVerse)
# Load example data 📊
data("count_matrices")
# Infer networks 🧠
networks <- infer_networks(
count_matrices_list = count_matrices,
method = "GENIE3",
nCores = 4
)
# Create consensus magic ✨
wadj <- generate_adjacency(networks)
consensus <- create_consensus(wadj, method = "vote")
# Visualize! 🎨
plotg(list(consensus))📚 Documentation
| Resource | Link | Description |
|---|---|---|
| 🌐 Website | ngsfc.github.io/scGraphVerse | Main documentation hub |
| 📖 Simulation Study | Vignette | Benchmarking tutorial |
| 🔬 Case Study | Vignette | Real-world example |
| 📋 Reference | Manual | Function documentation |
📝 Citation
citation("scGraphVerse") # 🎓 Academic credit🌟 Please also cite the original methods:
| Method | Citation | Journal |
|---|---|---|
| GENIE3 🌳 | Huynh-Thu et al. (2010) | PLOS ONE 5(9):e12776 |
| GRNBoost2 🚀 | Moerman et al. (2019) | Bioinformatics 35(12):2159-61 |
| ZILGM 🎯 | Park et al. (2021) | Statistical Analysis and Data Mining 37(18):3085-3092 |
| JRF 🌲 | Petralia et al. (2015) | Journal of Proteome Research 31(12):i197-i205 |
| PCzinb 🔗 | Nguyen et al. (2023) | Ann. Appl. Stat. 17(3):2555-73 |
| INet-Tool 🔧 | Policastro et al. (2025) | Comput Stat 40, 1517–1539 |
| Robin 🎯 | Policastro et al. (2021) | The R Journal 13(1):292-309 |
⚖️ License
scGraphVerse is licensed under GPL (≥ 2) 📜
🤝 Integrated Code Attribution
This package includes adapted functions from: - ZILGM (Park et al., 2021) - GPL-2 license - JRF (Petralia et al., 2015) - GPL (≥ 2) license - learn2count (Nguyen et al. 2023) - for the PCzinb implementation
All integrated code maintains proper attribution and copyright notices.