Beyond networks: Toward adaptive models of biological complexity

Area of Science:

• Neuroscience
• Network Science
• Computational Biology

Background:

• Network science models are valuable for complex systems, including neuroscience.
• Biological complexity presents unique modeling challenges beyond traditional network frameworks.

Purpose of the Study:

• To explore constructive enhancements for network models in biology.
• To address conceptual challenges like context dependence and history sensitivity in biological systems.
• To propose innovative modeling strategies that integrate dynamic and data-driven approaches.

Main Methods:

• Reviewing existing network science formalisms.
• Incorporating concepts like time-varying connections, adaptive topologies, and multilayer structures.
• Discussing biological concepts such as the "adjacent possible" and dynamic state spaces.
• Advocating for data-driven approaches to infer system properties.

Main Results:

• Identified opportunities to enhance network models by incorporating temporal dynamics and multilevel interactions.
• Highlighted the need for modeling strategies that account for biological context dependence, open-endedness, and history sensitivity.
• Emphasized the dynamic evolution of biological state spaces.

Conclusions:

• Network science can be fruitfully adapted to better capture biological complexity.
• Integrating dynamic, multilayer, and data-driven approaches is crucial for advancing biological modeling.
• Methodological and conceptual innovation is needed to deepen the explanatory power of network science in biology.