Computational protocol: Evolutionary history of the genus Tarentola (Gekkota: Phyllodactylidae) from the Mediterranean Basin, estimated using multilocus sequence data

Similar protocols

Protocol publication

[…] Tissue from tail tip muscle was collected from each individual and preserved in 96% ethanol. Genomic DNA was extracted using the DNeasy Extraction Kit from Qiagen following the manufacturer's protocol. A total of 384 individuals were used, belonging to all known species of Tarentola in the Mediterranean Basin. Some of the individuals included in this study were already used in previous works [,-,,]. Geographic location of each specimen is represented in Figure , and detailed information about locality and amplified genes in Table S1 from Additional file .For all individuals the Polymerase Chain Reaction (PCR) amplification and sequencing of two mtDNA gene fragments, the 12SrRNA and 16SrRNA was performed using the primers 12Sa/12Sb and 16Sar/16Sbr from Kocher et al. [] and Palumbi [], respectively. PCR conditions were the same as those described in Harris et al. []. Four nuclear protein-coding gene fragments were also sequenced: the acetylcholinergic receptor M4 (ACM4), the melanocortin 1 receptor (MC1R), the phosducin (PDC), and the recombination activating gene 2 (RAG2). For amplification and sequencing of ACM4 the primers tg-F and tg-R published by Gamble et al. [] were used. Regarding the MC1R fragment the primers MC1R_F and MC1R_R from Pinho et al. [] were used, and primers PHOF2 and PHOR1 [] for the amplification and sequencing of PDC. Amplification of ACM4, MC1R and PDC fragments were carried out in 25 μl volumes, containing 5.0 μl of 10x reaction Buffer, 2.0 mM of MgCl2, 0.5 mM each dNTP, 0.2 μM each primer, 1 U of Taq DNA polymerase (Invitrogen), and approximately 100 ng of template DNA. Finally, amplification and sequencing of the RAG2 gene fragment was performed using two sets of primers; 31FN.Venk/Lung.460R (amplification) and Lung.35F/Lung.320R (amplification and sequencing) published by Hoegg et al. []. PCR conditions were the same as described in Chiari et al. []. All amplified fragments were sequenced in a ABI3730XL automated sequencer.The obtained sequences were imported into the software Geneious Pro v5.4.0 [] where alignment was performed with MAFFT v6.814b [] using the default parameters (auto algorithm; scoring matrix = 200 PAM/k = 2; gap open penalty = 1.53; and offset value = 0.123). All sequences generated in this study were submitted to GenBank with accession numbers ranging from JQ300539 to JQ301443. Detailed information on the individuals and sequences are described in the Table S1 from Additional file . [...] Using the software ALTER [] the mtDNA dataset was reduced to unique haplotypes, considering gaps as differences. Regarding the nuclear loci, heterozygous positions were coded with the corresponding ambiguity letter.In order to determine the best fitting nucleotide model for each gene (mtDNA and nDNA), we used the software jModelTest v0.1.1 [], under the Akaike Information Criterion [following []]. Maximum Likelihood (ML), and Bayesian Inference (BI) phylogenetic analyses were performed for both the concatenated mitochondrial and nuclear DNA datasets, and for all the loci (mtDNA+nDNA). ML analyses were conducted with the software RAxML v7.2.8 alpha [], partitioning the dataset per locus. For all analyses 20 thorough ML searches were performed in order to obtain the best ML tree with support values, and thereafter 1000 bootstrap inferences. A majority rule consensus tree was generated using the software Phyutility []. BI was implemented with the program Mr.Bayes v3.1.2 [] under a partitioned model (dataset divided into genes), and considering the model of nucleotide substitution estimated with jModelTest. The Bayesian posterior probabilities were estimated using a Metropolis-coupled Markov chain Monte Carlo sampling approach, and both runs started with random trees running for 10 × 106 generations, saving one tree every 100 generations producing 100,000 trees. Both convergence and appropriate sampling were confirmed by examining the standard deviation of the split frequencies between the two simultaneous runs and the Potential Scale Reduction Factor (PSRF) diagnostic. The first 25,000 trees of each run were included in the burn-in period and discarded. Next, a majority-rule consensus tree was generated from the remaining trees. In both phylogenetic analyses Ptyodactylus hasselquistii was used as outgroup.The T. boettgeri group from the Canary and Selvages Islands has been used in previous analyses to provide a calibration point for estimates of the time of the cladogenetic events for the phylogeny of this genus [,], but the lack of available nuclear DNA sequences precludes its use in this study As an alternative, the substitution rate of the same mitochondrial region calculated for Tarentola was used for this purpose. Mean substitution rates and the standard error of the mean values for exactly the same 12S region as in the present study was extracted from a fully-calibrated phylogeny of Tarentola from the Canary islands [,]. This value was used as an informative prior in our divergence dating analysis. Specifically, we set a normal distribution prior for the ucld.mean parameter of the 12S partitions based on the result of the meanRate posterior (mean and standard error) of the calibration analyses of Tarentola (0.00891 ± 0.0000376 for the 12S).We used BEAST v.1.6.1 [] to estimate dates of the cladogenetic events from the concatenated dataset. The dataset comprised sequences from all six genes (nuclear genes unphased) using a phylogeny pruned arbitrarily to include one representative from each of the major lineages uncovered with the concatenated analysis (29 specimens in total). This method excludes closely related terminal taxa because the Yule tree prior (see below) does not include a model of coalescence, which can complicate rate estimation for closely related sequences []. Analyses were run four times for 5 × 107 generations with a sampling frequency of 10,000. Models and prior specifications applied were as follows (otherwise by default): GTR+I+G (12S, 16S), TN93 (ACM4), HKY+G (MC1R), HKY (PDC), TN93+G (RAG2); Relaxed Uncorrelated Lognormal Clock (estimate); Yule process of speciation; random starting tree; alpha Uniform (0, 10); yule.birthRate (0, 1000); ucld.mean of 12S Normal (initial value: 0.00827, mean: 0.00827, Stdev: 0.00162). Convergence for all model parameters was assessed by examining trace plots and histograms in Tracer v1.4 [] after obtaining an effective sample size (ESS) > 200. The initial 10% of samples were discarded as burn-in. Runs were combined using LogCombiner, and maximum credibility trees with divergence time means and 95% highest probability densities (HPDs) were produced using Tree Annotator (both part of the BEAST package). Trees were visualized using FigTree v1.3.1 (available at http://tree.bio.ed.ac.uk/software/figtree). […]

Pipeline specifications

Software tools MUSCLE, Geneious, MAFFT, jModelTest, RAxML, phyutility, BEAST, FigTree
Applications Phylogenetics, Nucleotide sequence alignment
Organisms Tarentola mauritanica