Characterizing Variability and Uncertainty Associated with Transcriptomic Dose-Response Modeling.

Area of Science:

• Environmental toxicology
• Genomics
• Risk assessment

Background:

• Transcriptomics dose-response analysis (TDRA) is a key tool for toxicogenomics and risk assessment.
• Understanding variability and uncertainty in TDRA experimental design is crucial but not well-established.

Purpose of the Study:

• To evaluate the impact of experimental design variations on TDRA results.
• To identify optimal experimental parameters for reducing uncertainty in transcriptomic point-of-departure (tPOD) calculations.

Main Methods:

• Analysis of 55 RNA-seq profiles from Japanese quail liver tissue exposed to chlorpyrifos.
• Subsampling the dataset 637 times to simulate various dose ranges, spacing, and replicate numbers (n=2-5).
• Calculation of gene and pathway benchmark doses and transcriptomic point-of-departure (tPOD) values using different methods.

Main Results:

• TDRA revealed substantial variability in gene and pathway benchmark doses across different designs.
• Overall tPOD values showed relative stability when calculated using "pathway" and "mode" methods.
• tPOD values were more sensitive to dose range and spacing than to the number of replicates.

Conclusions:

• Optimal TDRA experimental designs should prioritize a wider range and spacing of doses over a higher number of replicates.
• Fewer replicates (n=2 or 3) with more dose groups can reduce uncertainty in tPOD estimation.
• tPOD values can vary significantly (over tenfold) across designs, necessitating interpretation as order-of-magnitude estimates.