Deriving time-concordant event cascades from gene expression data: A case study for Drug-Induced Liver Injury (DILI)

Area of Science:

• Toxicology and Pharmacology
• Systems Biology
• Computational Biology

Background:

• Adverse event pathogenesis involves complex molecular and phenotypic changes.
• Adverse Outcome Pathways (AOPs) formalize these sequences using causal evidence, including temporal order.
• Understanding the temporal dynamics of toxicity is crucial for mechanistic insights.

Purpose of the Study:

• To investigate time-concordance of molecular events preceding adverse outcomes using a data-driven approach.
• To generate hypotheses on potentially causal mechanisms in toxicology.
• To explore the application of time-resolved transcriptomics within the AOP framework.

Main Methods:

• Analysis of time-resolved gene expression data from rat liver following repeat-dose studies (TG-GATEs database).
• Application of "first activation" concept to identify time-concordant events preceding histopathology (surrogate for Drug-Induced Liver Injury).
• Integration of transcription factor activity and prior knowledge on functional interactions to derive gene-regulatory mechanisms.

Main Results:

• Identified time-concordant gene expression events preceding DILI-related histopathology.
• Confirmed known DILI mechanisms and identified novel events, such as Sox13 activation.
• Demonstrated that significance, frequency, and log fold change can prioritize events.

Conclusions:

• Time-resolved transcriptomics and time-concordance analysis effectively support mechanistic hypothesis generation in toxicology.
• Quantifying temporal relationships aids in understanding toxicity mechanisms and advancing the AOP framework.
• Developed a Shiny app for interactive exploration of time-concordant events in DILI.