Figure 1

Coverage plots for example data sets. A and B) Artificial data sets where organisms are represented by dots with Euclidean distances. The concentric circles around the labelled organisms indicate how many organisms would be well represented by this choice of model strain for different threshold values. C and D) Coverage plots for the data sets in A and B respectively, showing how many organisms are well represented for increasing threshold values. Key for D: a – x, b – o, c – Δ. E) Best Total Coverage score (as a percentage of maximum) for 1–30 model organisms, for data set B, the model organisms are selected using a greedy implementation of the BTC method as described later in the text.