OpenTox 2022 Virtual Conference
TXG-MAPr tools: gene co-expression network analysis linked to histopathology provides quantitative mode-of-action assessment and prediction of drug-induced toxicity
In drug development it is crucial to detect or prevent unanticipated organ toxicity at an early stage. Integration of toxicogenomic data with histopathology is an useful approach to associate molecular and cellular mechanisms with pathogenesis. Therefore, we have applied weighted gene co-expression network analysis (WGCNA) on publicly available toxicogenomic rat liver and kidney TG-GATEs datasets and developed an interactive R-Shiny based TXG-MAPr tools for data visualization. Perturbation of WGCNA gene networks (modules) were quantitatively assessed by module eigengene scores (EGs) for each treatment condition, which Z-scores and summarizes the log2 fold change of all genes in a module. These module EGs were statistically associated with pathology phenotypes, providing prognostic information for drug safety assessment. In this way we could link pathogenesis to specific gene co-expression modules that represent cellular processes, including cell cycle, immune response, cell adhesion / cytoskeleton integrity, cell migration, extracellular matrix remodeling and RNA processing. Interestingly, modules that showed the highest statistical association with pathogenesis contain several well-known and novel biomarkers of kidney injury. Using preservation statistics we showed that several of these rat co-expression modules were also preserved in other test systems, like human tissue, indicating that pathogenic processes, including renal biomarker expression, could translate across species. Finally, new transcriptomic data can be uploaded in the TXG-MAPr tool for visualization of gene network perturbation and comparison with TG-GATEs datasets to identify underlying mechanisms of toxicity by chemical insults. In conclusion, the TXG-MAPr represents an innovative and powerful tool that contributes to drug safety assessment by providing mechanistic understanding of potential adverse drug reactions.
CV: Steven Kunnen did his bachelor and master Life Science & Technology at Delft University of Technology and Leiden University. After his graduation he did his PhD at the Leiden University Medical Center (LUMC) at the department of Human Genetics in the field of polycystic kidney disease. Steven continued his career to investigate drug induced kidney and liver injury as postdoctoral scientist at Leiden University. In this field he combined his molecular biological knowledge with bioinformatics by applying gene co-expression network analysis and building TXG- MAPr tools to investigate mechanisms of chemical induced toxicity.