Aligning Statistical Dynamics Captures Biological Network Functioning

TL;DR

This paper introduces a novel method for aligning biological networks to better infer their functions and predict dynamics, using statistical dynamics comparison across diverse biological systems.

Contribution

It proposes a new approach of aligning networks based on their statistical dynamics, enabling functional classification across various biological scales.

Findings

Successfully distinguished underlying processes in synthetic and real-world networks.

Improved inference of network functions across biological systems.

Demonstrated applicability from cellular to ecosystem levels.

Abstract

Empirical studies of graphs have contributed enormously to our understanding of complex systems. Known today as network science, what was originally a theoretical study of graphs has grown into a more scientific exploration of communities spanning the physical, biological, and social. However, as the quantity and types of networks have grown so has their heterogeneity in quality and specificity. This has hampered efforts to develop general network theory capable of inferring functioning and predicting dynamics across study systems. We have successfully approached this challenge by aligning networks to each other rather than comparing parameter estimates from individually fitted models or properties of edge topologies. By comparing the predictability of statistical dynamics originating from each network's constituent nodes we were able to build a functional classifier that distinguished underlying processes in both synthetic and real-world network data spanning the entire biological scale from cellular machinery to ecosystems.