Recursive model for dose-time responses in pharmacological studies.

Area of Science:

• Pharmacology
• Biostatistics
• Computational Biology

Background:

• Clinical studies frequently monitor dose-response curves over time.
• Existing models like the Hill equation capture dose-response at a single time point, while Gompertz equation models temporal changes at a fixed dose, but not both simultaneously.
• There is a need for models that integrate both dose and time dynamics for comprehensive response prediction.

Purpose of the Study:

• To propose a novel parametric model that integrates dose-response dynamics (Hill equation) with temporal evolution (Gompertz law).
• To develop a regression framework linking model parameters to individual-level proteomic data for personalized predictions.
• To enable accurate prediction of dose-response curves over time for new individuals.

Main Methods:

• Developed a parametric model combining Gompertz law for temporal changes and Hill equation for dose-dependency.
• Derived a recursion relation to capture the temporal evolution of dose-response curves.
• Specified a regression model to connect dose-time response parameters with individual proteomic data.

Main Results:

• The proposed Recursive Hybrid model effectively captures both dose-response relationships and their temporal evolution.
• The joint model successfully predicts dose-response curves over time for new individuals based on their proteomic data.
• Demonstrated superior performance of the Recursive Hybrid model compared to individual predictive models using synthetic and pharmacological data.

Conclusions:

• The Recursive Hybrid model offers improved accuracy in predicting dose-time varying drug responses.
• The model's effectiveness was validated using the HMS-LINCS database for anticancer compounds.
• This approach facilitates genetic characterization and personalized drug response prediction.