SEEDS: data driven inference of structural model errors and unknown inputs for dynamic systems biology

Area of Science:

• Systems biology
• Computational biology
• Bioinformatics

Background:

• Dynamic models, often ordinary differential equations, are crucial for understanding biological systems.
• Challenges in mechanistic modeling include inaccurate knowledge of interactions and unknown environmental inputs.
• Effective interventions and experimental design rely on accurate biological models.

Purpose of the Study:

• To introduce the R-package SEEDS, which implements algorithms for inferring structural model errors and unknown inputs.
• To aid researchers in addressing misfits between models and experimental data.
• To enhance the efficiency of model recalibration and experimental design.

Main Methods:

• The SEEDS R-package utilizes two novel algorithms.
• These algorithms infer structural model errors from output measurements.
• Unknown environmental inputs are also identified using these algorithms.

Main Results:

• SEEDS facilitates the identification of structural discrepancies in dynamic biological models.
• The package helps uncover unmeasured inputs affecting biological systems.
• This leads to more accurate mechanistic models and better-informed experimental designs.

Conclusions:

• The SEEDS R-package provides essential tools for improving mechanistic modeling in biology.
• By addressing model errors and unknown inputs, SEEDS enhances the reliability of biological system analysis.
• This supports more effective targeted interventions and future research directions.