Fig. 4: Performance of TMarSel-selected markers on MAGs from the EMP dataset.

a Topological accuracy, measured as the proportion of confident branches, of trees inferred from 793 complete MAGs from the EMP dataset. Left panels indicate the tree quality of previous marker sets, whereas middle and right panels show the tree quality of KEGG and EggNOG markers, respectively. Symbols indicate trees inferred from different marker sets (color-coded), while line style and filled or unfilled symbols distinguish trees inferred from MAGs containing at least 25% of markers versus those containing only one marker. b Mean topological accuracy of trees inferred from simulations of increasingly 793 incomplete MAGs across n = 10 replicates. c nRF distance between trees inferred from incomplete MAGs to complete MAGs. The incompleteness threshold refers to the proportion of ORFs removed from each MAG. Unfilled bars indicate that MAGs contained at least one marker. Data are presented as the mean ± standard deviations. Asterisks show the significance level of two-sided two-sample Mann-Whitney U tests after Benjamini–Hochberg correction between the quality of trees inferred from KEGG versus PhyloPhlAn3 markers. Significance levels: n. s.: p > 0.05; *: p ≤ 0.05; **: p ≤ 0.01; ***: p ≤ 0.001. Source data and test statistics are provided as a Source Data file.