Fig. 8
From: Skmer: assembly-free and alignment-free sample identification using genome skims
The mean rank and distance error of the best remaining match in leave-out experiments. The distance of the closest genome in the reference to a query is varied from 0.01 to 0.1 (x-axis). The rank and distance errors (y-axis) of the best match to a query, are computed by comparing the order given by each method with the order obtained by applying Mash* to the full assemblies (ground truth). For each dataset, the experiment is repeated by taking each species as the query, and then the errors are averaged. Three methods, Mash, Skmer, and AAF, are compared on a the Anopheles dataset, b the Drosophila dataset, and c the avian dataset