GrowthPredict: A toolbox and tutorial-based primer for fitting and forecasting growth trajectories using...

Area of Science:

• Mathematical Biology
• Applied Statistics
• Epidemiology

Background:

• Accurate short-term forecasting of growth processes, such as disease outbreaks, requires accessible dynamic modeling tools with quantified uncertainty.
• Existing user-friendly toolboxes for real-time forecasting of time-series trajectories using phenomenological growth models are limited.

Purpose of the Study:

• Introduce and illustrate GrowthPredict, a MATLAB toolbox for fitting and forecasting time-series data using ordinary differential equation-based growth models.
• Provide a user-friendly resource for students and researchers in mathematical biology, applied statistics, and infectious disease modeling.

Main Methods:

• Developed GrowthPredict, a MATLAB toolbox implementing phenomenological dynamic growth models (exponential, generalized growth, Gompertz, generalized logistic, Richards).
• Incorporated functions for time-series forecasting, uncertainty quantification via parametric bootstrapping, and performance assessment across various models and data conditions.
• Utilized publicly available data, including the monkeypox (mpox) epidemic in the USA, for demonstration and validation.

Main Results:

• GrowthPredict offers a flexible and accessible platform for fitting and forecasting diverse growth trajectories.
• The toolbox enables the generation of real-time short-term forecasts with quantified uncertainty, crucial for decision-making.
• Demonstrated the utility of GrowthPredict through examples and a tutorial video, highlighting its application in disease outbreak analysis.

Conclusions:

• GrowthPredict provides a valuable, user-friendly resource for characterizing and forecasting time-series data using simple dynamic growth models.
• The toolbox facilitates informed policy decisions regarding control strategies and intervention impact assessment for contagion processes.
• Enhances the ability to conduct real-time, uncertainty-quantified forecasts for natural and societal growth phenomena.