Performance of ascertainment schemes explored across 27 exhaustively sampled <i>f</i>_<i>4</i>-statistic classes and assessed as residual standard deviation of a linear trend fitted to a scatterplot of Z-scores on unascertained vs. ascertained data (abbreviated as “residual SE” and expressed in the same units as <i>f</i>_<i>4</i>-statistic Z-scores).

Only f4-statistics having absolute Z-scores <15 SE (on all sites) were considered. We also applied the binary classifier to determine if an ascertainment produces unbiased or biased results (the latter cases are highlighted in bold and underlined text). The numbers of f4-statistic classes or ascertainment schemes affected by bias are shown in the rightmost column and in the bottom row, respectively. For each f4-statistic class, total number of distinct statistics and the number of statistics having absolute Z-scores <15 SE (on all sites) are shown. The composition of the f4-statistic classes is shown above the table in an abbreviated way: AFR, African populations; AMRSIB, Native American and Siberian populations; ARCH, archaic human individuals (Neanderthals and Denisovans); chimp, chimpanzee; EAS, East Asian populations; EUR, European populations; ME, Mediterranean and Middle Eastern populations; PAP, Papuan and Australian populations. The SNP counts correspond to sites polymorphic in larger collections of groups from which the analyzed population quintuplets were taken, see S2 Table. SNP counts vary across the population sets, and minimal and maximal values are shown in separate columns. (XLSX)