TinderMIX

TinderMIX models integrated time- and dose-dependent gene expression to identify dynamic dose-response behaviors in toxicogenomics studies.

Key Features:

• Simultaneous Modeling: Fits integrated time- and dose-response models to each gene to capture combined temporal and dosage effects.
• Dynamic Dose-Response Analysis: Identifies genes exhibiting time-dependent dose responses by comparing fits of various integrated models.
• Optimal Model Selection: Selects the best-fitting model for each gene to represent its response pattern accurately.
• Time and Dose Effect Mapping: Computes maps of time and dose effects for genes to represent response magnitude across conditions.
• Responsive Area Identification: Applies a user-defined threshold to effect maps to pinpoint areas of significant dynamic response and label responsive genes.

Scientific Applications:

• Identify Dynamic Mechanisms: Determine how drugs or chemicals affect molecular pathways over time and across dose levels using gene-level time-dose responses.
• Compare Toxicity Profiles: Analyze and compare mechanisms and toxicity potential of compounds, as demonstrated with cyclosporin A and thioacetamide in the Open TG-GATEs dataset.

Methodology:

Start from gene log fold-change data, fit various integrated time-dose models to each gene, select the best model per gene, compute time-dose effect maps, apply a user-defined threshold to identify responsive areas, and label responsive genes based on integrated time and dose points of departure.