A network of paths toward innovation

Abstract

Department of Biology, University of Pennsylvania, Philadelphia, PA, USA E-mail: jdraghi@gmail.com Any mention of neutral change and adaptation in the same breath inevitably brings to mind the battles over the causes of molecular evolution that have been so prominent and long-running in the field. While these arguments have evolved with the development of new models and new technologies, the central dichotomy has remained: beneficial changes, that fix because of natural selection and contribute to adaptation, are one thing, and neutral substitutions, that fix only by chance, are something quite different. Wagner has been a cogent and prolific advocate of a radical alternative: neutral variation within a population can provide an essential foundation for adaptive change. In experimental papers [1], numerous simulation studies, reviews [2], and his previous book [3], Wagner has emphasized how epistasis – interactions among genes that shape the adaptive consequences of mutations – allows neutral variation to modulate the effects of other mutations, producing a vast range of possible mutant phenotypes that may lead to an adaptive innovation. Wagner did not invent the idea of neutral variation fueling adaptation, but he has done much of the hard work to translate the idea into themainstream of biology. The early foundations of this idea came from the pioneering work of investigators who had trained as physicists and chemists, and applied biophysical ideas and computer models of macromolecules to evolutionary questions [4–8]. In his 2005 book Evolvability and Robustness in Living Systems, Wagner synthesized these studies with other evidence for neutral networks across a broad range of scales, drawing together the genetic code, macromolecular structures, regulatory and metabolic networks, and developmental systems. The magnitude of this evidence for degeneracy and epistasis, and of its implications for evolution, was exciting and influenced a broad audience. But Wagner stopped short of presenting a fully realized theory of how all this potential for neutral variation could explain the astounding ability of life to produce true innovation in the face of environmental change and opportunity. In this new book, Wagner moves his focus from robustness to ‘‘innovability’’ – an evolutionary propensity to innovate – but retains the central theme of evolution across networks of connected genotypes. Rather than building to a dramatic unveiling of his theory of innovation, Wagner delivers his main argument after less than a third of the book; the remaining chapters each develop the theory around a specific concern, such as recombination or plasticity. This structure recalls Darwin’s Origin of Species, in which the theory of natural selection was laid out by chapter four and then applied for the remaining 11 chapters. In retrospect, we can see the tremendous potential latent in those few simple ingredients that make up the idea of natural selection. But is Wagner’s theory of the origins of innovation rich enough to warrant its own book? To answer this question, we have to appreciate the scope of Wagner’s goals. While Wagner sets out his explicit goals in an ambitious first chapter, the next three chapters give the first real glimpses of what this book is meant to accomplish. Wagner devotes 50 pages to a quantitative picture of the genotype networks that underlie metabolic systems, transciptionally regulated gene networks, and protein and RNA biomolecules. Drawing extensively on published work from the Wagner lab, these chapters establish the foundation of DOI 10.1002/bies.201200016