Spontaneous mutations emerge as drivers of de novo gene birth in a minimal evolutionary model

For long it was believed that new genes only arise from old genes. This view was challenged by the discovery of orphan genes, which show no homology with any conserved, established genes. In fact, genomics data suggests that de novo gene birth, where new genes are born from previously non-genic sequences, is ubiquitous and frequent. Nevertheless, we still lack a general mechanism for the process of gene birth, and therefore have no basis to form expectations for the extent and timescale of the process. We propose here that spontaneous mutations are potential drivers of gene birth. We use a simple evolutionary model and ask whether accumulating beneficial mutations are sufficient to turn a non-genic locus into a de novo gene. We sampled hundreds of biologically reasonable distributions of fitness effects (DFE) of mutations, and found that gene birth occurred for a majority of these DFEs. This is broadly in line with observations from genomics, suggesting that spontaneous mutations on their own could indeed serve as a general mechanism of de novo gene birth.