Optimal designs for the individual and joint exposure general logistic regression models

Area of Science:

• Pharmacology and Toxicology
• Biostatistics
• Experimental Design

Background:

• Research on combined compound administration aims to improve efficacy and reduce adverse effects.
• Existing statistical work primarily focuses on modeling joint dose-response curves, with less attention to experimental design for joint action assessment.
• Parametric dose-response models, often nonlinear, are common in this field.

Purpose of the Study:

• To develop a methodology for designing experiments to assess the joint action of compounds.
• To provide a flexible approach applicable to various response types and error structures.
• To enable designs that are compound-independent by focusing on proportionate responses.

Main Methods:

• Utilized a nonlinear weighted least squares approach for experimental design.
• Adapted methods for continuous and discrete responses, alternative error structures, and nonzero background responses.
• Focused on additive and nonadditive independent joint action (IJA) models.

Main Results:

• Developed a nonlinear weighted least squares method for experimental design in joint action studies.
• Proposed designs expressed in terms of proportionate responses, enhancing applicability across different compounds.
• Extended previous work on optimal and minimal experimental designs for single compounds.

Conclusions:

• The proposed nonlinear weighted least squares methodology offers a robust framework for designing joint action experiments.
• The approach is versatile, accommodating various statistical models and data types.
• This work provides a foundation for more efficient and informative experimental designs in combination toxicology and pharmacology.