Figure 2. PhyloTempo graphical output summarizing phylogenetic tree shape statistics. Note: A tree from the OPTIONS data set (patient p5, unique sequences) with a high TC statistic (0.7) was used.

Summary:

Two new measures of tree topology are introduced: temporal clustering (TC), and staircase-ness. Several other existing statistics are also implemented for the purpose of comparison: Aldous’s graphical test and likelihood test to decide if a tree fits the Yule or uniform model; Coless’s and Sackin’s shape statisitics under both Yule and uniform hypotheses; cherry count; and Pybus’s gamma.

R libraries used are: “ape” (Analysis of Phylogenetics and Evolution); “ade4” (Analyse de Données destinée d’abord à la manipulation des données Écologiques et Environnementales avec des procédures Exploratoires d’essence Euclidienne); “phybase” (phylogenetic trees – removed from CRAN); “phylobase” (base package for phylogenetic trees); “phangorn” (phylogenetic analysis using maximum likelihood, maximum parsimony, distance methods, and Hadamard conjugation); “doBy” (do something on data which is grouped by some variables); “infotheo” (information theory); “apTreeshape” (simulation and analysis of tree topologies using statistical indices); and “diversitree” (diversification and character evolution).

Abstract:

Serially-sampled nucleotide sequences can be used to infer demographic history of evolving viral populations. The shape of a phylogenetic tree often reflects the interplay between evolutionary and ecological processes. Several approaches exist to analyze the topology and traits of a phylogenetic tree, by means of tree balance, branching patterns and comparative properties. The temporal clustering (TC) statistic is a new topological measure, based on ancestral character reconstruction, which characterizes the temporal structure of a phylogeny. Here, PhyloTempo is the first implementation of the TC in the R language, integrating several other topological measures in a user-friendly graphical framework. The comparison of the TC statistic with other measures provides multifaceted insights on the dynamic processes shaping the evolution of pathogenic viruses. The features and applicability of PhyloTempo were tested on serially-sampled intra-host human and simian immunodeficiency virus population data sets. PhyloTempo is distributed under the GNU general public license at https://sourceforge.net/projects/phylotempo/.

Source:

Melissa M. Norström, Mattia C.F. Prosperi, Rebecca R. Gray, Annika C. Karlsson, Marco Salemi, PhyloTempo: A Set of R Scripts for Assessing and Visualizing Temporal Clustering in Genealogies Inferred from Serially Sampled Viral Sequences. Evol Bioinform Online. 2012; 8: 261–269. Published online 2012 June 11. doi: 10.4137/EBO.S9738. PMCID: PMC3382462.