From graph topology to ODE models for gene regulatory networks

Area of Science:

• Systems Biology
• Computational Biology
• Bioinformatics

Background:

• Gene regulatory networks (GRNs) are modeled using directed graphs and ordinary differential equations (ODEs).
• Qualitative graph models represent causal gene interactions, while quantitative ODE models track molecular concentrations over time.
• A clear link between these two modeling approaches is needed.

Purpose of the Study:

• To establish a formal connection between graph-based and ODE-based models of gene regulatory networks.
• To introduce and define the 'constant sign property' for ODE models.
• To investigate the conditions under which ODE models are consistent with signed, directed graphs.

Main Methods:

• Proposed the 'constant sign property' for ODE models.
• Analyzed global and local conditions for ODE-graph consistency.
• Examined ODE models for cis-regulatory elements and protein complex binding.
• Investigated model fitting between different ODE models.

Main Results:

• The constant sign property precisely defines the conditions for ODE-graph consistency.
• Global constant sign property implies a unique graph for an ODE model.
• Local constant sign property means an ODE model can correspond to different graphs under varying conditions.
• ODE models of cis-regulatory elements with protein binding satisfy the global constant sign property.
• Multiple ODE models can be consistent with a single gene regulatory graph.
• Model fit is better when ODE models originate from the same graph.

Conclusions:

• The constant sign property provides a rigorous framework for linking ODE and graph models of GRNs.
• Understanding this connection is crucial for accurate systems biology modeling.
• The study offers insights into the complexity and parameterization of gene regulatory models.