Most Networks in Wagner's Model Are Cycling

摘要: In this paper we study a model of gene networks introduced by Andreas Wagner in the 1990s that has been used extensively to evolution mutational robustness. We investigate range features and parameters evaluate extent which they influence probability random network will produce fixed point steady state expression pattern. There are many different types models literature, (discrete/continuous, sparse/dense, small/large network) attempt put some order into diversity, motivated fact properties qualitatively same all models. Our main result is give rise cyclic behavior more often than points. And although periodic orbits seem dominate dynamics, usually considered unstable not allowed survive previous evolutionary studies. Defining stability as points, show distribution these highly robust changes its parameters. also find sparser be stable, may help explain why favored evolution. have unified several disconnected studies class under framework stability, way had systematically explored before.