Identification of neutral biochemical network models from time series data

Area of Science:

• Systems Biology
• Computational Biology
• Biophysics

Background:

• Modeling biological systems from time series data is challenging due to difficulties in parameter estimation and system structure identification.
• Canonical models, adhering to strict guidelines, can simplify these modeling tasks.

Purpose of the Study:

• To propose a method for identifying admissible parameter sets of canonical S-systems from biological time series.
• To explore parameter estimation and network topology identification in biological system modeling.

Main Methods:

• A Monte Carlo process combined with an improved parameter optimization algorithm was used.
• The method maps parameter space to network space by creating an ensemble of decoupled S-system models.
• The concept of 'sloppiness' was revisited to explore parameter sets and network topologies yielding similar dynamics.

Main Results:

• The methodology was applied to time series data from the glycolytic pathway of Lactococcus lactis, yielding ensembles of models with varying network topologies.
• Comparison with a pre-specified topology approach highlighted the flexibility of the proposed method.
• The study demonstrated that different parameter sets and network topologies can produce similar dynamical behaviors.

Conclusions:

• The proposed parameter estimation method serves as a powerful tool for hypothesis testing in systems biology.
• It facilitates the design of new experiments by exploring potential biological system structures and regulations.
• The findings contribute to a deeper understanding of biological pathway dynamics and model identifiability.