TY - JOUR T1 - The Underlying Molecular and Network Level Mechanisms in the Evolution of Robustness in Gene Regulatory Networks A1 - Pujato, Mario A1 - MacCarthy, Thomas A1 - Fiser, Andras A1 - Bergman, Aviv Y1 - 2013/01/03 N2 - Author Summary Development from egg to embryo depends to a large extent on regulatory networks of genes called transcription factors. Previous research has shown these gene regulatory networks to be robust to perturbations at the level of the connections between transcription factors. Here, we investigate the mechanisms underlying the evolution of robustness in gene networks using a modeling approach, which considers three levels: binding of individual transcription factors to DNA, dynamics of gene expression levels, and fitness effects at the population level. In our model the gene regulatory network is determined by transcription factor binding sites within DNA sequences, which undergo mutation. We categorize these mutations in a continuum ranging from silent mutations, which have no effect on regulation and change only the DNA sequence (local-sequence level), to mutations that change connections between genes in the network (network-architecture level). We find that in sparse networks, containing few connections between genes, a balance of local-sequence and network-architecture level mechanisms are responsible for the evolution of robustness, but when the network is densely connected the network-architecture level mechanisms become dominant. We argue that the shift towards the network-architecture level for more densely-connected networks offers a potential explanation for the evolution of increased complexity. JF - PLOS Computational Biology JA - PLOS Computational Biology VL - 9 IS - 1 UR - https://doi.org/10.1371/journal.pcbi.1002865 SP - e1002865 EP - PB - Public Library of Science M3 - doi:10.1371/journal.pcbi.1002865 ER -