Quartet Consistency Count Method for Reconstructing Phylogenetic Trees

Abstract

Among the distance based algorithms in phylogenetic tree reconstruction, the neighbor-joining algorithm has been a widely used and effective method. We propose a new algorithm which counts the number of consistent quartets for cherry picking with tie breaking. We show that the success rate of the new algorithm is almost equal to that of neighbor-joining. This gives an explanation of the qualitative nature of neighborjoining and that of dissimilarity maps from DNA sequence data. Moreover, the new algorithm always reconstructs correct trees from quartet consistent dissimilarity maps.2000 Mathematics Subject Classification. 92D15, 68R10, 05C05, 68Q25.

Constructing Semi-Directed Level-1 Phylogenetic Networks from Quarnets

“…Phylogenetic trees can be constructed using cherry-picking methods to peel off tree cherries and recursively build up trees [2]. Networks can be constructed analogously by peeling off exterior structures.…”

“…The first method is based on sequentially adding leaves to an existing network and can be viewed as a generalization of quartet-puzzling [15]. The second algorithm, which we call the cherry-blob algorithm is a generalization of cherry picking on trees [2], where we recursively identify the external structures of a network which can include both tree cherries and also cycles in a level-1 network.…”