Fig. 2.

Performance of the tools under three different scenarios. The performance of tools supporting replicates (left panels) and not supporting replicates (right panels) was analyzed separately. For fast tools, the entire data set was analyzed (solid line) whereas a subset of the data was used for slow tools (dashed lines); The performance of a random classifier is shown as the reference (black dotted line). a Selective sweeps. b Truncating selection. c Stabilizing selection