Abstract
This chapter surveys recent developments in simulating the evolution of GRNs in developmental biology. Over the past two decades, computational biologists have developed a number of approaches to study how developmental GRNs evolve. This has led to a number of breakthroughs in understanding the mechanisms of how species maintain their body plans, and how they evolve or speciate in response to environmental perturbations. EA uses the general evolutionary processes of repeated mutation, reproduction and selection in optimization problems. The progress in computational biology described here has deepened and refined understanding of the biological principles underlying these processes. Our aim is for this chapter to provide some inspiration to computer scientists in EA to incorporate new biologically inspired techniques. We feel this offers a large potential for improving EA efficiency. In turn, computational biology could greatly benefit from EA research, for instance in multi-objective optimization, coding of multiscale problems, and efficiencies in solution techniques. Following a brief survey of the major trends in the computational biology approaches, we discuss the refinements these have made to understanding evolutionary mechanisms. In particular, we discuss the factors affecting GRN evolvability and robustness; the effect different genetic alteration mechanisms (e.g. types of mutation) have on evolutionary speed and robustness; the role of network growth; modelling co-evolution; modelling multi-factor control of gene expression; and applying these techniques to the evolution of GRNs controlling spatially-dependent gene expression (underlying embryonic tissue differentiation). We finish with a brief summary of how these might be incorporated into and improve EA searches.,Book chapter,Published.