GAN-WGCNA: Calculating gene modules to identify key intermediate regulators in cocaine addiction

Area of Science:

• Genomics and Systems Biology
• Neuroscience and Behavioral Biology

Background:

• Understanding gene expression dynamics over time is crucial for disease diagnosis and treatment.
• Next-generation sequencing (NGS) data offers insights into biological regulatory mechanisms.
• Generative adversarial networks (GANs) can augment biological data to reveal hidden gene expression profiles.

Purpose of the Study:

• To develop a pipeline for analyzing augmented time-series gene expression data.
• To create an unbiased map of biological processes and identify intermediate regulators.
• To investigate gene expression patterns during cocaine addiction using advanced analytical methods.

Main Methods:

• Utilized a pipeline integrating GANs with weighted gene co-expression network analysis (GAN-WGCNA) and rescued differentially expressed gene (rDEG) methods.
• Analyzed transcriptome data from mice undergoing cocaine self-administration.
• Interpreted spatiotemporal gene expression information and screened for intermediate genes.

Main Results:

• Identified two significant genes, Alcam and Celf4, as missed intermediate regulators correlated with cocaine addiction behavior.
• Demonstrated the statistical significance of these genes' correlation with addiction.
• Mapped the spatiotemporal expression and co-regulation of Alcam and Celf4 across time, brain regions, and biological processes.

Conclusions:

• The developed pipeline effectively identifies key intermediate genes in complex biological processes like addiction.
• Alcam and Celf4 are highlighted as potential therapeutic targets for cocaine addiction.
• GAN-WGCNA provides a powerful tool for visualizing and analyzing time-series gene module interplay and phenotype correlations.