Limits to Predicting Evolution: Insights from a Long-Term Experiment with Escherichia coli

Abstract

Our inability to predict how populations of cells will evolve is a fundamental challenge to human health and biological engineering. In medicine, one would like to predict and thwart, or at least have time to adequately prepare for potentially catastrophic events such as the emergence of new pathogens, the spread of drug resistance, and the progression of chronic infections and cancers. In bioengineering, one would like to stop, or at least delay, evolution that inactivates a designed function, in order to make genetic engineering and synthetic biology more reliable and efficient. On a larger scale, one would also like to predict when the presence of recombinant DNA or a certain species might pose a threat to nature or civilization if it has the potential to evolve to become harmful.