Computer model replicates complex networks behind evolution

You inhabit something of a miracle, in engineering terms. Your body consists of trillions of cells, woven together into something whose complexity far outstrips that of the most sophisticated objects our best engineers can produce, from computers and skyscrapers to space shuttles.

A relatively simple outer form belies a teeming society of chemical reactions and protein engineering. This must maintain itself within strict temperature and physiological limits while enduring a complex and frequently unpredictable external environment. And, to achieve its long lifespan, it must avoid the sort of catastrophic breakdown that plagues human-engineered objects.

All the breathtaking innovation required to produce this complexity rests on two pillars of evolution that are, for the most part, either ignored or unappreciated. These are robustness and evolvability, which together grant what evolutionary biologist Andreas Wagner calls “innovability” in his engaging and intelligent Arrival of the Fittest. Wagner’s message is that these two foundation stones of evolution exist because of an unexpected and remarkable degree of neighbourliness (not his term) that seems to characterize life — a neighbourliness that allows species to innovate more rapidly and successfully than previously imagined.

To get from simple replicating molecules through to single-celled organisms such as bacteria and eventually on to complex and ungainly multicellular organisms like giant squid, natural selection has had to search through a vast library of varieties and combinations of genes. Now, imagine you are in the squid section of the library and you want to make an albatross. Every step along the way has to be something that works: it has to be a competitive organism.

Wagner has discovered what makes this search possible. It is good neighbours, and lots of them. The genes that make our bodies typically do not act alone. Instead, they form large and complex networks that interact to produce metabolisms, tissues and organs. He has built computer models of these networks in which he randomly alters some feature, mimicking in silico the sort of random mutation that natural selection relies on. He then asks whether the mutated network as a whole can still perform the job it was designed to do.

Leave a Reply

glp menu logo outlined

Newsletter Subscription

* indicates required
Email Lists
glp menu logo outlined

Get news on human & agricultural genetics and biotechnology delivered to your inbox.