Computational protocol: Contrasting patterns of Andean diversification among three diverse clades of Neotropical clearwing butterflies

Similar protocols

Protocol publication

[…] We aligned sequences with CodonCode Aligner v6.0.2. The molecular dataset was partitioned by gene and codon positions. We performed a maximum‐likelihood analysis including all individuals using IQ‐TREE software as implemented in the W‐ID‐TREE server (Nguyen, Schmidt, von Haeseler, & Minh, ; Trifinopoulos, Nguyen, von Haeseler, & Minh, ). IQ‐TREE automatically selected the best partition scheme, and we performed 1000 ultrafast bootstrap analyses (Minh, Nguyen, & von Haeseler, ). The tree can be found in Appendix . [...] We computed a consensus sequence for each species and gene region, resulting in a dataset containing 86 species and 44 outgroups. We used the “greedy” algorithm and linked rates implemented in PartitionFinder 1.1.1 (Lanfear, Calcott, Ho, & Guindon, ) to select the best models of substitution for optimized sets of nucleotides over all models implemented in BEAST. A time‐calibrated phylogeny was generated using BEAST 1.8.2 (Drummond, Suchard, Xie, & Rambaut, ), using the best partition scheme and uncorrelated lognormal relaxed clock for each partition. We applied eight secondary calibration points based on Nymphalidae node ages obtained from Wahlberg et al. () and Solanaceae host‐plant ages following De‐Silva et al. () (Appendix ). We used conservative uniform priors for secondary calibrations, with upper and lower bounds corresponding to those of the 95% credibility intervals inferred in Wahlberg et al. (), or to the upper (more ancient) bound of the Solanaceae lineage age inferred by Magallón, Gómez‐Acevedo, Sánchez‐Reyes, and Hernández‐Hernández () and De‐Silva et al. () and zero (present), because host‐plant calibrations are maximum calibrations. Each run was performed for 100 million generations and sampled every 100,000 generations, resulting in 1,000 trees. The maximum clade credibility tree using the median of posterior distribution for node ages was extracted using TreeAnnotator (Drummond et al., ), applying a 10% burn‐in (Appendix ‐). We also ran two independent analyses to assess the effect on node ages of using a Birth‐Death tree prior or a Yule prior. Results were extremely similar (Appendix ), and we used the Birth‐Death tree for all analyses. […]

Pipeline specifications

Software tools CodonCode Aligner, IQ-TREE, PartitionFinder, BEAST
Application Phylogenetics
Organisms cellular organisms