Regulatory networks and connected components of the neutral space

TL;DR

This study explores the structure and fragmentation of the neutral space of regulatory networks in yeast, revealing how mutations affect network properties and the proximity of the wildtype to optimal configurations.

Contribution

It analyzes the neutral space of yeast regulatory networks, showing its fragmentation and the properties of networks within the wildtype component compared to others.

Findings

The neutral space is fragmented into approximately 470 million components.

Networks in the wildtype component are sparser and more noise resilient.

Networks reachable from wildtype have fewer fixed point attractors.

Abstract

The functioning of a living cell is largely determined by the structure of its regulatory network, comprising non-linear interactions between regulatory genes. An important factor for the stability and evolvability of such regulatory systems is neutrality - typically a large number of alternative network structures give rise to the necessary dynamics. Here we study the discretized regulatory dynamics of the yeast cell cycle [Li et al., PNAS, 2004] and the set of networks capable of reproducing it, which we call functional. Among these, the empirical yeast wildtype network is close to optimal with respect to sparse wiring. Under point mutations, which establish or delete single interactions, the neutral space of functional networks is fragmented into 4.7 * 10^8 components. One of the smaller ones contains the wildtype network. On average, functional networks reachable from the wildtype by mutations are sparser, have higher noise resilience and fewer fixed point attractors as compared with networks outside of this wildtype component.