Why does evolution work? Randomly switching 0s and 1s in the assembly language of a computer program will essentially never improve its function. Nevertheless, a seemingly similar mutation process in DNA gives rise to the remarkable diversity of life we see around us. Why does one work so well and the other so badly? This question of evolvability is central to our work (Masel & Trotter 2010).
We study evolvability using theoretical models, especially those that explicitly capture mechanistic constraints, usually from biochemistry or genetics, and calculate their evolutionary consequences. We also use existing genome data to bioinformatically test hypotheses that come out of our theories.
The partial robustness of biological systems is important to their evolvability. Most biological variants either break a component, or tinker with it: their effects are rarely in between. Adaptation always comes from tinkering mutations, never from breaking: there is no environment in which lethal mutations are adaptive. When the phenotypic effects of mutations are partially suppressed, selection remains strong enough to weed out the most deleterious mutations, enriching the pool of cryptic variants, by a process of elimination, for those most likely to contribute to evolvability (Masel 2006; Rajon & Masel 2011). This process may be important in the origin of de novo coding sequences from noncoding ancestors (Giacomelli et al. 2007; Wilson & Masel 2011). We also study “evolutionary capacitors” such as the yeast prion [PSI+], which seems to be an adaptation to tap into and exploit such pre-adapted stocks of variation (Masel & Bergman 2003; Griswold & Masel 2009; Lancaster et al. 2010).
We study a variety of other topics in theoretical population genetics, including the molecular clock (Peterson & Masel 2009), bet-hedging (King & Masel 2007), mutation accumulation (Masel et al. 2007; Maughan et al. 2007), genetic drift (Masel 2012), and evolutionary rescue in the context of density-dependent hard selection and the substitutional load.