Fig. 5

The likelihood for a sequence to have a neighbor (one point mutant) of a degeneracy k filtered by the degeneracy of the sequence itself. It is most likely for a sequence of degeneracy 2,3,4 or 6 to have a neighbor of the same degeneracy while for degeneracy 5, 7 (and higher odd numbers) it’s most likely to have a non-degenerate neighbor. We see that neighbors of degenerate sequences are far more likely to be degenerate compared to neighbors of nondegenerate sequences. This implies that degenerate sequences tend to form networks in sequence space