4.2. Sequence Analysis, Alignment, and Phylogenetic Analysis

New sequence contigs were assembled using Geneious V. 11 (Biomatters Ltd., New Zealand). Sequences were compiled and cleaned to keep just the open reading frame. Nucleotide sequences were aligned using the online version of MAFFT v.7 [89]. The alignment was refined by hand, using Mesquite V. 3.5 [90], considering protein domains and amino acid motifs that have been reported as conserved for KNOX genes. A matrix that included KNOX, the ELK, and the TALE-HD was used for phylogenetic analyses. Phylogenetic relationships were inferred from the nucleotide data using Bayesian inference (BI). Analyses were performed on CIPRES (http://www.phylo.org) [91]. The best partition scheme was found with PartitionFinder2 [92], for the nucleotides matrix 15 data blocks were defined by dividing the matrix into five regions (KNOX 1 (first KNOX domain), KNOX 1–KNOX 2 (region between KNOX 1 and KNOX 2), KNOX 2 (second KNOX domain), KNOX 2–HD (region between KNOX 2 and the HD), and HD (ELK and TALE-HD domains)), and by dividing each region by codon position. Analyses were performed with nine subsets as estimated by the corrected Akaike Information Criterion (AICc) implemented in PartitionFinder2 (Table S1; see Supplemental Data with this article). For the amino acids matrix, the JTT+I+G model was used as estimated by the corrected Akaike Information Criterion (AICc) implemented in PartitionFinder2 [93]. For both matrices, BI analyses were conducted using MrBAYES 3.2.6 [93]. Two independent runs of 10 million generations were completed, with four chains each (three heated, one cold), using a chain temperature of 0.2 and uniform priors. Trees and parameters were sampled every 1000th generation. Samples corresponding to the initial phase of the Markov chains (25%) were discarded as burn-in. The applicability of this burn-in value was determined by the inspection of the likelihood scores and effective sample sizes. Post-burn-in trees were combined to obtain a single majority rule consensus tree and the respective posterior probabilities (PPs) of nodes. Trees were depicted using FigTree v1.4.3 (http://tree.bio.ed.ac.uk/software/figtree/).