Constructing graphs from genetic encodings

Area of Science:

• Network Science
• Computational Biology
• Systems Biology

Background:

• Real-world connected systems evolve through various mechanisms, including random and growth-dependent topologies.
• The role of innate, identity-dependent compatibility rules in network formation has been historically overlooked.
• Understanding these rules is crucial for modeling complex systems like brains, social networks, and biological interactions.

Purpose of the Study:

• To derive a network encoding process based on genetic principles that allows for the reproducible generation of specific network topologies.
• To explore the representational power of this approach by generating diverse network structures.
• To introduce a novel null-model for analyzing identity-driven connectivity in biological and social systems.

Main Methods:

• Development of a network encoding process inspired by genetic principles and brain connectivity.
• Proposal of stochastic and deterministic processes for generating a wide range of network topologies.
• Characterization of a Random Genetic (RG) family of networks and comparison with Erdős-Rényi graphs.

Main Results:

• Demonstrated ability to reproducibly generate specific network topologies using identity-dependent wiring rules.
• Successfully generated structured graphs, including feed-forward and hierarchical networks.
• Characterized the Random Genetic (RG) family, showing critical phase transitions and distinct behaviors under targeted attacks due to modular underpinnings.

Conclusions:

• The proposed framework provides a powerful method for generating and analyzing network topologies based on innate compatibility rules.
• This approach offers a relevant null-model for social and biological systems where identity influences connectivity.
• The study highlights the importance of identity-dependent rules in shaping network structure and function.