Fig. 4: Human variation at fourfold degenerate sites is negatively associated with phyloP score.
From: Interpreting mammalian synonymous site conservation in light of the unwanted transcript hypothesis
a The proportion of 4d sites that are variable (minor allele frequency >0; n = 716,661) based on single-nucleotide polymorphism data from TOPMed data freeze 8 was calculated in phyloP bins. For all four bases, the proportion of variable sites is significantly negatively correlated with mean phyloP in each bin (two-tailed Pearson’s r = −0.86, −0.91, −0.74, and −0.80, P = 0.012, 0.0048, 0.022, and 0.010 for A, T, C and G bases, respectively). b Same as (a) but considering only variants with a minor allele frequency >0.001 (n = 27,818). Again, all four bases show a significant negative correlation with mean phyloP in each bin (two-tailed Pearson’s r = −0.94, −0.92, −0.98, and −0.97, P = 0.0019, 0.0038,4.9 × 10−6, and 1.04 × 10−4 for A, T, C and G bases, respectively). c When considering only significantly conserved 4d sites (phyloP > = 2.27; n = 544,503), we observed a significant difference in the proportions of variant types compared to expected proportions based on total numbers of each variant type independent of phyloP (two-tailed Pearson’s Chi2 test; χ2 = 261, d.f. = 11, P < 2.2 × 10−16). d Residuals from the Chi2 test show that there is a deficiency in variable A and T sites and an excess of variable C and G sites amongst conserved 4d sites. MAF minor allele frequency. Source data are provided as a Source Data file.
