Fig. 4

Realignment of included and held-out assemblies to the pangenomes. a Edit rate of the twelve assemblies used in pangenome construction. Each faded dot represents one autosome of each assembly in either the bulk or centro-/telomeric ranges. Mean and 95% confidence intervals are also indicated for each category of sequence per assembly. b Similar plot to a but for eight additional assemblies held-out from pangenome construction. c Query coverage for the bulk sequence for each autosome of all assembly. Values above the dashed line indicate there were more aligned bases than query bases, suggesting multiple ambiguous alignments that were equally scored. d Similar to c, but for the centro-/telomeric sequence. e Bar plots of for the number of alignment errors reported by GraphAligner when using finite “tangle effort”. Bar heights reflect the mean across 348 (included) and 232 (not included) autosomes across the assemblies, while error bars represent 95% confidence intervals