(a), Plot from figure 2b; RelCP vs. functional ORGs (log10). Blue circles: species with low coverage (≤ 6x) genome assemblies (r2 = 0.76, P < 0.0001). (b), Adjusted regression plot. Number of functional ORGs in low-coverage genomes is increased by 39.7%. Percent increase is derived by comparing the number of ORGs extracted from earlier, low-coverage draft genomes to the number from current ≥ 6x coverage assemblies, averaged over four available relevant cases (elephant, dog, cat, rabbit). Adjusted correlation: r2 = 0.84, P < 0.0001. Any increase in ORG count in low-coverage species strengthens the correlation, as all but two have relatively few ORGs for their RelCP. (c), Omitting species with ≤6x genomes strengthens the correlation between RelCP and functional ORGs further (r2 = 0.89, P < 0.0001).

(a), Inversion of the axes from figure 3a showing that species with largest RelCP tend to have higher numbers of OR pseudogenes. (b), A similar non-linear, dual pattern emerges when the dependent variable in figure 3a is changed from RelCP to number of functional ORG. Species with > ~500 pseudogenes tend to have the largest functional ORG repertoires.