Fig. 4: Theoretical predictions and empirical observations of allelic heterogeneity.
a Predicted values of the number of independent variants of the same allele observed at fixation (k) as a function of mutation rates (u). Equation based on Haldane (1927) and Kimura (1962) and taken from Wilson et al. (2014). Predictions are based on an effective population size (Ne) of 50,000 and selection coefficient (s) of 0.01. Highlighted are frequently observed ranges of empirical estimates of mutation rates from classic population genetics (Muller 1928; Haldane 1933; Rhoades 1941; Stadler 1946, 1948) and sequence-based mutation rates from modern molecular genomics (Lynch et al. 2016). Inset figure illustrates the hypothetical dynamics of multiple independent alleles (each a different color) with positive selection. Collectively the variants increase in frequency, ultimately leading to fixation of adaptive variants (elimination of deleterious ancestral allele), but individually each variant remains at low frequency. b Detected levels of allelic heterogeneity in genes enriched for loss of function in humans (obs > exp) reported by (Karczewski et al. 2020). Highlighted are cases of previously studied genes with evidence of beneficial effects or positive selection CCR5 (Libert et al. 1998), SLC30A8 (Flannick et al. 2014; Dwivedi et al. 2019), TMPRSS11B (Updegraff et al. 2018), TRIM65 (Wang et al. 2016; Wei et al. 2018), PLA2G7 (Song et al. 2012), HDAC10 (Dahiya et al. 2020), CD36 (Fry et al. 2009; Love-Gregory et al. 2011).
