Examples of multree reduction in single locus data sets.

<p>A. Highly ranked tree for query “<i>Brassica</i>” inferred from a single-locus data set aligned at the level of Brassicaceae (data set #56065; phenylalanine ammonia-lyase). B. Reduction of the tree in A to species-level relationships without conflict. Note the loss of one taxon, <i>Brassica napus</i>; this taxon was present in many quartets in the original tree, but each one was contradicted by at least one other quartet. Therefore no conflict-free information was present for this taxon, and it was removed by the algorithm. C. Low ranked tree from query “<i>Drosophila</i>” (data set #130188; SMOX gene) in which there are no informative edges so the multree reduction produces a null result (no edges, no taxa; not shown). D. Intermediate ranked tree in which only one taxon has multiple sequences and the reduced singly-labeled tree contains all taxa (data set #91190; ‘yolk protein 1’; reduced tree not shown, but can be obtained by deleting all but one of the leaves labeled <i>Drosophila grimshawi</i>).</p>