Figure 6

Influence of genetic drift on the response to selection. Strong positive (synergistic) epistasis is often supposed to lead to sigmoid responses to selection, with an acceleration of the selection response when favorable genotypic combinations appear in the population. This is also what can be expected for the genetic network described in this article (see, e.g., Figure 3B). However, this sigmoid shape was not observed during the experimental selection process (see Figure 1). Among other hypotheses, random genetic drift can easily hide this theoretical acceleration of the selection response. This figure details 10 simulations of the first generations of the selection response in the high line (same parameters as Figure 3B), and shows how, despite a clear overall sigmoid shape, some single simulations can easily appear to respond very regularly to selection (one of them is highlighted).