CALUX measurements: statistical inferences for the dose-response curve.

Area of Science:

• Environmental Chemistry
• Toxicology
• Biotechnology

Background:

• The Chemical Activated LUciferase gene eXpression [CALUX] bioassay is crucial for detecting dioxin-like compounds.
• Standard CALUX analysis uses non-linear regression for quantifying 2,3,7,8-tetrachlorodibenzo-p-dioxin [TCDD] and determining Bioanalytical EQuivalents [BEQs].
• Assessing the uncertainty of predicted BEQ values presents a significant statistical challenge.

Purpose of the Study:

• To present a novel statistical approach using linear calibration functions with Box-Cox transformations for CALUX assay data.
• To address challenges in uncertainty assessment for predicted BEQ values.
• To compare statistical methods for optimizing BEQ determination and enhancing assay performance.

Main Methods:

• Application of Box-Cox transformations to linearize CALUX dose-response curves.
• Calculation of confidence and prediction intervals for CALUX response and predicted BEQ values.
• Statistical comparison of sigmoid and linearized models, including inverse prediction and effective concentration ratios (ECR(20-50-80)).

Main Results:

• Linear calibration functions with Box-Cox transformations effectively address BEQ uncertainty assessment.
• The study provides methods for calculating confidence and prediction intervals for both CALUX response and BEQ values.
• Statistical comparisons offer guidance for optimizing BEQ determinations using the H1L6.1c3 cell line.

Conclusions:

• Box-Cox transformations offer a robust statistical framework for improving the CALUX assay's reliability and uncertainty estimation.
• The proposed methods enhance the detection and estimation capabilities of the CALUX assay for dioxin-like compounds.
• This approach provides valuable guidance for accurate BEQ determinations in toxicological assessments.