Topological augmentation to infer hidden processes in biological systems

Area of Science:

• Systems Biology
• Computational Biology
• Biochemical Engineering

Background:

• Inferring the structure (topology) of biochemical systems is a significant challenge.
• System topology involves identifying all relevant molecules and their interactions.
• Current methods often lack statistical rigor and systematic approaches.

Purpose of the Study:

• To present a statistically rigorous and systematic method, topological augmentation, for inferring biochemical system structures.
• To integrate prior knowledge and experimental data for accurate model building.
• To provide a versatile tool applicable to various systems described by ordinary differential equations.

Main Methods:

• Topological augmentation iteratively refines models by adding terms to explain experimental data.
• Stochastic differential equations represent model topology uncertainty, guiding the augmentation process.
• The method is semi-automatic and guided by parameter space sampling.

Main Results:

• Successfully applied to a pharmacokinetic model, highlighting the importance of global parameter sampling.
• Improved understanding of yeast glutamine transport, revealing two distinct permease kinetics.
• Demonstrated applicability to diverse systems modeled by ordinary differential equations.

Conclusions:

• Topological augmentation provides a robust framework for uncovering complex system topologies.
• The method enhances the accuracy and completeness of biochemical system models.
• This approach facilitates deeper insights into biological mechanisms and dynamics.