The path towards prediction

Abstract

All living systems possess the ability to evolve. This ability has allowed life to adopt a large diversity of shapes, colors, sizes and lifestyles. However, despite being a fundamental property of life, our knowledge of what makes living systems evolvable is limited. The fact that evolution can be sometimes be hopelessly unpredictable while in other cases it follows only a small number of predictable paths has puzzled evolutionary biologists for decades. An important step towards resolving this problem has been the discovery that non-additive effects between mutations, a phenomenon coined epistasis, can act as a source of evolutionary constraint. Importantly, the limited number of accessible adaptive pathways in conditions where constraints are pervasive provides opportunities for prediction. In this thesis, we explore how these constraints can arise from the biochemical interactions within the cell. Aside from answering fundamental questions in biology, the ability to predict evolution will open up unprecedented possibilities, ranging from the engineering of living systems to the development of novel treatments for disease....