2.2. Phylogenetic Analyses
2.2.1. Parsimony Analyses
We identified MP trees in PAUP* 4.0b10 [45] using the parsimony ratchet [46]. Ratchet searches reweight a random subset of characters and conduct searches using those perturbed matrices, permitting a more thorough exploration of treespace (for a detailed explanation see Nixon [46]). For this study, the ratchet analyses used 100 iterations with 20% of informative characters perturbed and one tree held per iteration. To conduct the ratchet analyses, we used a C++ program (written by E.L.B.) that generates an appropriate PAUP* block. After conducting 100 ratchet iterations, the optimal trees were retained and tree bisection, and reconnection (TBR) branch swapping was conducted to identify the full set of MP trees. When we compared this strategy to a more typical tree search (random additions of taxa followed by TBR branch swapping), we found that the ratchet took a shorter amount of time and identified shorter trees. Ratchet bootstrap analysis used 500 replicates, each of which used 100 ratchet iterations, as described above, with the final swapping limited instead to 1,000 trees per bootstrap replicate.
2.2. Phylogenetic Analyses 2.2.1. Parsimony Analyses We identified MP trees in PAUP* 4.0b10 [45] using the parsimony ratchet [46]. Ratchet searches reweight a random subset of characters and conduct searches using those perturbed matrices, permitting a more thorough exploration of treespace (for a detailed explanation see Nixon [46]). For this study, the ratchet analyses used 100 iterations with 20% of informative characters perturbed and one tree held per iteration. To conduct the ratchet analyses, we used a C++ program (written by E.L.B.) that generates an appropriate PAUP* block. After conducting 100 ratchet iterations, the optimal trees were retained and tree bisection, and reconnection (TBR) branch swapping was conducted to identify the full set of MP trees. When we compared this strategy to a more typical tree search (random additions of taxa followed by TBR branch swapping), we found that the ratchet took a shorter amount of time and identified shorter trees. Ratchet bootstrap analysis used 500 replicates, each of which used 100 ratchet iterations, as described above, with the final swapping limited instead to 1,000 trees per bootstrap replicate.
การแปล กรุณารอสักครู่..
