Understanding Weighted Gene Co-Expression Network Analysis (WGCNA) with a Simple Example

Created in July 24, 2024

2024   ·   Bioinformatics   ·   Networks

Introduction

Weighted Gene Co-Expression Network Analysis (WGCNA) is a widely-used method in bioinformatics for identifying clusters (modules) of highly correlated genes. It is particularly useful for understanding complex biological networks and identifying key regulatory genes (hub genes) in various biological processes and diseases. In this post, we will delve into the step-by-step process of WGCNA with illustrative examples.

Steps in WGCNA

1. Filtering Genes

Before performing WGCNA, genes with low variance or incomplete data are filtered out to ensure the analysis focuses on the most informative genes.

Example:

Sample Gene1 Gene2 Gene3 Gene4 Gene5 Gene6
A 5 0 50 100 2 10
B 6 0 52 95 3 12
C 5 1 48 98 2 9
D 7 0 51 102 4 11
E 6 0 49 97 3 10

Genes with low variance (Gene2) and incomplete data (Gene5) are filtered out:

Sample Gene1 Gene3 Gene4 Gene6
A 5 50 100 10
B 6 52 95 12
C 5 48 98 9
D 7 51 102 11
E 6 49 97 10

2. Normalization

Data normalization helps in reducing non-biological variations. Methods like Z-score normalization are commonly used.

Example:

Sample Gene1 Gene3 Gene4 Gene6
A -1.0 -0.1 0.5 -0.5
B 0.0 1.3 -1.1 1.5
C -1.0 -1.5 -0.3 -1.5
D 1.0 0.7 1.1 0.5
E 0.0 -0.3 -0.5 0.0

3. Calculating Similarity Matrix

The similarity between pairs of genes is calculated using Pearson correlation.

Example:

Gene1 Gene3 Gene4 Gene6
Gene1 1.0 0.8 0.4 0.6
Gene3 0.8 1.0 0.5 0.7
Gene4 0.4 0.5 1.0 0.3
Gene6 0.6 0.7 0.3 1.0

4. Creating Weighted Adjacency Matrix

The similarity measures are raised to a power β to emphasize strong correlations and diminish weak ones.

Example:

Gene1 Gene3 Gene4 Gene6
Gene1 1.0 0.2621 0.0041 0.0467
Gene3 0.2621 1.0 0.0156 0.1176
Gene4 0.0041 0.0156 1.0 0.0007
Gene6 0.0467 0.1176 0.0007 1.0

5. Constructing Gene Network

Using the weighted adjacency matrix, a gene network is created where nodes represent genes, and edges represent weighted connections.

6. Identifying Modules

Clustering algorithms are used to identify modules, which are groups of genes with similar expression patterns.

Example Modules:

  • Module 1: Gene1, Gene3, Gene6
  • Module 2: Gene4

Conclusion

WGCNA is a powerful tool for identifying gene modules and understanding gene networks in large-scale biological data. This method helps in uncovering key regulatory genes and pathways involved in various biological processes and diseases, thereby providing insights that can drive future research and therapeutic strategies.

References

  1. Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008.
  2. Zhang B, Horvath S. A general framework for weighted gene co-expression network analysis. Stat Appl Genet Mol Biol. 2005.