S15 Fig -

Scatterplots illustrating the effects of 10 ascertainment schemes on Z-scores (a) or f4-statistics (b) of two classes: 854,358 distinct statistics including up to three African groups and/or up to four East Asian groups, pooled with 749,700 distinct statistics including American/Siberian and/or European and/or Papuan groups. Instead of individual points, heatmaps illustrating point density are shown. Z-scores on all sites (from ca. 5.3 to 15 million sites, as indicated on the x-axes) are compared to Z-scores on ascertained datasets on the y-axes. Ascertainment schemes and site counts are shown in plot titles. The 2nd and 4th rows of plots in each panel represent close-up views on the origin of the plots: they are based on statistics with absolute Z-scores below 15 on all sites. Linear trends fitted to the data and lines representing ± 2 SE are shown in red. Squared Pearson correlation coefficients (R2) for these data are shown on the y-axes. First, linear trends were fitted to the sets of Z-scores or f4-statistics lying near the origin (having absolute Z-score below 15 on all sites), and then the same linear equations were re-fitted to complete sets of Z-scores or f4-statistics. Since we show both f4-statistics and Z-scores side by side here, R2 was used instead of residual SE as a measure of correlation. (PDF)