TripNet: A triplet based approach for detecting reticulateevolution

The problem of constructing an optimal rooted phylogenetic network from a set of rooted triplets is an NP-hard problem. This paper presents a new approach for detecting reticulate evolution and reconstructing evolutionary histories in the presence of reticulation. Indeed we use TripNet, a new algorithm which reconstruct rooted phylogenetic networks from triplets, to infer explicit phylogenetic networks from (biological) data. TripNet is a heuristic method for constructing explicit rooted phylogenetic networks from triplets. To demonstrate the accuracy and efficiency of TripNet, we performed four simulations. The results show that TripNet constructs rooted phylogenetic networks near to the optimum and TripNet networks can reasonably show recombination between sequences. Moreover the results show that TripNet is a valuable method for removing noise in the data. Also we applied the method to five published data sets: Kreitman’s data, a set of triplets from real yeast data obtained from the Fungal Biodiversity Center in Utrecht, a collection of 110 highly recombinant Salmonella multi-locus sequence typing sequences, and nrDNA ITS and cpDNA JSA sequence data of New Zealand alpine buttercups of Ranunculus sect. Pseudadonis. Finally, we compare our results with those already obtained by other authors using alternative methods. The results show that TripNet can produce the same major clusters which are obtained from other methods.