Increasing Data Transparency and Estimating Phylogenetic Uncertainty in Supertrees: Approaches Using Nonparametric Bootstrapping

Abstract

The estimation of ever larger phylogenies requires consideration of alternative inference strategies, including divide-and-conquer approaches that decompose the global inference problem to a set of smaller, more manageable component problems. A prominent locus of research in this area is the development of supertree methods, which estimate a composite tree by combining a set of partially overlapping component topologies. Although promising, the use of component tree topologies as the primary data dissociates supertrees from complexities within the underling character data and complicates the evaluation of phylogenetic uncertainty. We address these issues by exploring three approaches that variously incorporate nonparametric bootstrapping into a common supertree estimation algorithm (matrix representation with parsimony, although any algorithm might be used), including bootstrap-weighting, source-tree bootstrapping, and hierarchical bootstrapping. We illustrate these procedures by means of hypothetical and empirical examples. Our preliminary experiments suggest that these methods have the potential to improve the correspondence of supertree estimates to those derived from simultaneous analysis of the combined data and to allow uncertainty in supertree topologies to be quantified. The ability to increase the transparency of supertrees to the underlying character data has several practical implications and sheds new light on an old debate. These methods have been implemented in the freely available program, tREeBOOT.

Untitled

“…rQS avoids the problems that affect QS identified by Wilkinson et al [69]. Other supertree-specific metrics for assessing support, such as V [69], triplet-based methods [70] and modified bootstrap methods [71,72], have also been recently proposed, but are not used here. All rQS values were determined using the Perl script QualiTree [39].…”

A likelihood look at the supermatrix–supertree controversy

“…The supermatrix-supertree debate does not appear to have appreciated the fact that statistical likelihood is a natural tool for combining information from heterogeneous datasets and for dealing with problems such as missing data, non-random taxa sampling and non-overlapping taxa sets. Recently a few attempts have been made to develop bootstrap algorithms to assess and incorporate uncertainties in the subtrees in supertree construction, using parsimony for tree reconstruction (Burleigh et al, 2006;Moore et al, 2006). Some insights may be gained into this problem from a statistical modeling framework, by examining how the likelihood function incorporates information and accommodates uncertainties in multiple heterogeneous datasets.…”

“…Most supertree methods use heuristic algorithms that cannot be justified rigorously on a statistical basis. They also ignore uncertainties in the estimated subtrees (such as bootstrap support values, Bayesian posterior clade probabilities, or estimated branch lengths), although heuristic algorithms are recently proposed to remedy this problem (Burleigh et al, 2006;Moore et al, 2006).…”

A supertree of Temnospondyli: cladogenetic patterns in the most species-rich group of early tetrapods

“…from complete support to total conflict. Our dataset is not amenable to input tree bootstrapping (Creevey et al 2004; see also Burleigh et al 2006) due to sparse taxon sampling (Moore et al 2006). Wilkinson et al (2005b) have shown that biases occur in most supertree methods, and within the limits of feasibility, method dependence of results should be assessed.…”