Performance of Flip Supertree Construction with a Heuristic Algorithm

Abstract

Supertree methods are used to assemble separate phylogenetic trees with shared taxa into larger trees (supertrees) in an effort to construct more comprehensive phylogenetic hypotheses. In spite of much recent interest in supertrees, there are still few methods for supertree construction. The flip supertree problem is an error correction approach that seeks to find a minimum number of changes (flips) to the matrix representation of the set of input trees to resolve their incompatibilities. A previous flip supertree algorithm was limited to finding exact solutions and was only feasible for small input trees. We developed a heuristic algorithm for the flip supertree problem suitable for much larger input trees. We used a series of 48- and 96-taxon simulations to compare supertrees constructed with the flip supertree heuristic algorithm with supertrees constructed using other approaches, including MinCut (MC), modified MC (MMC), and matrix representation with parsimony (MRP). Flip supertrees are generally far more accurate than supertrees constructed using MC or MMC algorithms and are at least as accurate as supertrees built with MRP. The flip supertree method is therefore a viable alternative to other supertree methods when the number of taxa is large.

A likelihood look at the supermatrix–supertree controversy

“…Some methods are also suggested to work specifically with the distance or Bayesian methods of phylogeny reconstruction (Criscuolo et al, 2006;Ronquist et al, 2004). The different supertree methods appear to have very different statistical properties and their relative performance is a focus of much recent research (e.g., Goloboff and Pol, 2002;Pisani and Wilkinson, 2002;Gatesy et al, 2004;Bininda-Emonds, 2004b;Eulenstein et al, 2004;Wilkinson et al, 2005;Goloboff, 2005). However, all of them appear to lack a rigorous statistical justification and fail to account for uncertainties in the estimated subtrees.…”

“…However, all of them appear to lack a rigorous statistical justification and fail to account for uncertainties in the estimated subtrees. For example, Eulenstein et al (2004) compared several supertree algorithms in a simulation study. The different loci had the same sequence length, the same rate and the same number of species, so that the information content is about the same at all loci.…”

Exact ILP solutions for phylogenetic minimum flip problems

“…While the MFCT problem is fixed-parameter tractable in the optimal number of flips [1,11], for the MFST there exists only a slow (feasible up to 20 taxa in practice) exact method [2] and several heuristics [8,3], of which [3] empirically represents the state of the art. To investigate the practicality of the engineered Mfg approach, we conduct a set of experiments on a regular PC (Intel Core-i7 2.67GHz).…”

Many hexapod groups originated earlier and withstood extinction events better than previously realized: inferences from supertrees

“…As the best-characterized and most widely used method, standard MRP (Baum 1992;Ragan 1992) was used for a series of initial analyses, in which the dataset was refined (see dataset refinement in the electronic supplementary material). Five other supertree methods were also implemented: Purvis MRP (Purvis 1995), matrix representation with compatibility (Ross & Rodrigo 2004), matrix representation with flipping (Eulenstein et al 2004), the most similar supertree method or distance fit (Creevey et al (2004), and the average consensus (Lapointe & Cucumel 1997). We avoided using agreement supertree methods (Bininda-Emonds 2004) because they only identify relationships common to all input trees, providing little information when the many competing hypotheses of hexapod phylogeny are taken into account.…”