Technical Comments

Comment on "Hexapod Origins: Monophyletic or Paraphyletic?"

See allHide authors and affiliations

Science  12 Sep 2003:
Vol. 301, Issue 5639, pp. 1482
DOI: 10.1126/science.1086558

Figures

  • Fig. 1.

    Bayesian 50% majority rule consensus tree with associated branch lengths obtained using nucleotide sequences of cox1, cox2, cox3, and cytb (3750 sites) corresponding to the 35-taxon data set in (1). The first and third codon positions were RY-coded, whereas second codon positions were kept as nucleotides. MrBayes version 3.0b4 (12) was used to perform a partitioned-likelihood Bayesian search in which three independent substitution models were attributed to each codon position—a two-state substitution model + I + Γ for RY-coded first and third codon positions, and a GTR + I + model for second codon position nucleotides. Four incrementally heated Metropolis-coupled Markov chain Monte Carlo (MCMCMC) were run for 500,000 generations, sampling trees and parameters every 10 generations. The consensus tree was obtained from the 35,000 trees sampled after the initial burn-in period. Values at nodes indicate Bayesian posterior probabilities (* = 1.00). Note that the terminal branch lengths leading to the bee (Apis) and louse (Heterodoxus) have been reduced by a factor of three. Underlined taxa are not included in the 25-taxon data set.

  • Fig. 2.

    Maximum likelihood (ML) phylogram obtained using nucleotide sequences of cox1, cox2, cox3, and cytb for a 25-taxon data set (3777 sites). The third codon positions were RY-coded, whereas first and second codon positions were kept as nucleotides. PAUP* (13) was used to perform a ML heuristic search under the best fitting GTR + I + Γ model and associated ML estimates of parameters as determined by Modeltest version 3.06 (14). A partitioned-likelihood Bayesian search was carried out with MrBayes (12) using a GTR + I + Γ model for first and second codon position nucleotides and a two-state substitution model + I + Γ for the RY-coded third codon positions, with the same parameter settings as in Fig. 1. Values at nodes indicate ML bootstrap proportions (100 replications)/Bayesian posterior probabilities. The two collembolans are figured in bold.

Stay Connected to Science

Navigate This Article