high dimensional WGCNA

hdWGCNA is an R package for performing weighted gene co-expression network analysis (WGCNA) in high dimensional transcriptomics data such as single-cell RNA-seq or spatial transcriptomics. hdWGCNA is highly modular and can construct context-specific co-expression networks across cellular and spatial hierarchies. hdWGNCA identifies modules of highly co-expressed genes and provides context for these modules via statistical testing and biological knowledge sources. hdWGCNA uses datasets formatted as Seurat objects. Check out the hdWGCNA in single-cell data tutorial or the hdWGCNA in spatial transcriptomics data tutorial to get started.

Seurat v5 compatibility As of hdWGCNA v0.3.00 Seurat version 5 is now supported in addition to Seurat v4.

If you use hdWGCNA in your research, please cite the following papers in addition to the original WGCNA publication:

• Morabito et al., Cell Reports Methods (2023)
• Morabito & Miyoshi et al., Nature Genetics (2021)

Installation

We recommend creating an R conda environment environment for hdWGCNA.

# create new conda environment for R
conda create -n hdWGCNA -c conda-forge r-base r-essentials

# activate conda environment
conda activate hdWGCNA

Next open R and install the required dependencies:

• Bioconductor, an R-based software ecosystem for bioinformatics and biostatistics.
• Seurat, a general-purpose toolkit for single-cell data science.
• WGCNA, a package for co-expression network analysis.
• igraph, a package for general network analysis and visualization.
• devtools, a package for package development in R.

# install BiocManager
install.packages("BiocManager")

# install Bioconductor core packages
BiocManager::install()

# install additional packages:
BiocManager::install(c("WGCNA", "igraph", "devtools", "GeneOverlap", "ggrepel", "UCell"))
devtools::install_github("NightingaleHealth/ggforestplot")

# install Seurat v5 
install.packages("Seurat")

# alternatively, install Seurat v4
install.packages("Seurat", repos = c("https://satijalab.r-universe.dev', 'https://cloud.r-project.org"))

Now you can install the hdWGCNA package using devtools.

devtools::install_github('smorabit/hdWGCNA', ref='dev')

Suggested Reading

Check out the paper describing hdWGCNA, and our original description of applying WGCNA to single-nucleus RNA-seq data:

• hdWGCNA identifies co-expression networks in high-dimensional transcriptomics data

• Single-nucleus chromatin accessibility and transcriptomic characterization of Alzheimer’s disease

For additional reading, we suggest the original WGCNA publication and papers describing relevant algorithms for co-expression network analysis:

• WGCNA: an R package for weighted correlation network analysis
• Defining clusters from a hierarchical cluster tree: the Dynamic Tree Cut package for R
• Eigengene networks for studying the relationships between co-expression modules
• Geometric Interpretation of Gene Coexpression Network Analysis
• Is My Network Module Preserved and Reproducible?

Note about package development: hdWGCNA is under active development, so you may run into errors and small typos. We welcome users to write GitHub issues to report bugs, ask for help, and to request potential enhancements.