Sensitivity analysis of biological Boolean networks using information fusion based on nonadditive set functions

Area of Science:

• Systems Biology
• Network Science
• Computational Biology

Background:

• Boolean networks are crucial for modeling biological systems.
• Assessing network sensitivity to node mutations is vital for understanding system behavior.
• Nonadditive set functions offer a framework for information fusion in complex networks.

Purpose of the Study:

• To develop and apply an algebraic method for information fusion to assess Boolean network attribute contributions.
• To evaluate the joint impact of node attributes on network sensitivity to mutations.
• To classify attribute contributions in a biochemical signal transduction network.

Main Methods:

• Utilized an algebraic method based on nonadditive set functions for information fusion.
• Assessed node attributes including in-degree, out-degree, path lengths, bias, average sensitivity, and canalizing degrees.
• Measured mutation impact using average Hamming distance between wild-type and mutated networks.

Main Results:

• The algebraic method robustly classified attribute contributions in a biochemical signal transduction network.
• Out-degree and average sensitivity were identified as the most significant attributes influencing network sensitivity.
• The study focused on a network of signaling molecules, primarily proteins.

Conclusions:

• Both topological and dynamical properties of network nodes are important for analysis.
• The algebraic information fusion method is robust to initial conditions and data partitioning.
• This approach provides a reliable way to understand attribute contributions in complex biological networks.