On modeling and state estimation for genetic regulatory networks with polytopic uncertainties

Area of Science:

• Systems Biology
• Computational Biology
• Bioinformatics

Background:

• Gene expression data offers insights into organismal structures and behavior.
• Modeling gene regulatory networks (GRNs) from experimental data is a growing research area.
• Uncertainty quantification and state estimation are key challenges in GRN modeling.

Purpose of the Study:

• To investigate uncertainty quantification and state estimation for GRN models.
• To develop methods for handling high-dimensional gene expression data.
• To design robust and efficient state estimators for GRNs.

Main Methods:

• Utilized polytopic uncertainty models (PUMs) to represent parameter uncertainties in GRNs.
• Proposed the principal component plane (PCP) algorithm to reduce polytope complexity for high-dimensional models.
• Developed a system equivalence transformation for simplifying GRN models.
• Employed semi-definite programming to incorporate robust stability and H2 performance for state estimation.

Main Results:

• The PCP algorithm effectively prunes polytopes while retaining essential information.
• System equivalence transformation simplifies the GRN model for state estimation.
• A novel vertex-dependent condition for H2 estimators reduces conservatism.
• Simulations on real-world microarray data demonstrate the effectiveness of the H2 estimators.

Conclusions:

• The proposed methods provide a robust framework for uncertainty quantification and state estimation in GRNs.
• The H2 estimators are capable of handling short, high-dimensional gene expression time series.
• This work contributes to a better understanding of gene regulatory mechanisms through data-driven modeling.