Analysis of gene sequence similarity and phylogeny Sequence data were edited and assembled in Omiga 2.0 and EMBOSS GUI (European Molecular Biology Open Software Suite [56] and gene alignments were manually checked and optimized using BioEdit v.7.0.9
[57] and MEGA 4 [58]. GC content and the location of polymorphic sites were analyzed using Omiga 2.0 and FaBOX [59] (http://www.birc.au.dk/software/fabox). All seven CUDC-907 genes (flaA, recA, pyrH, ppnK, dnaN, era, and radC) were concatenated using Se-Al ver.2.0a11 [60], giving a final alignment of 6,780 nucleotides (including gaps). The range of intraspecific sequence similarity (%) for each gene was calculated using the sequence identity matrix program implemented in BioEdit. Nucleotide polymorphism in each gene was evaluated by quantifying the nucleotide SGC-CBP30 in vivo diversity per site (Pi) using DNA Sequence Polymorphism software (DnaSP 5.10) [61].
Maximum Likelihood (ML) and Bayesian methods were used to analyze both individual genes, and concatenated gene sequence datasets. The optimal substitution model and gamma rate heterogeneity for Cilengitide in vitro individual genes and combined dataset were determined using the Akaike Information Criterion (AIC) in MrModeltest ver. 2.2 [62]. Maximum likelihood (ML) trees were generated using GARLI ver. 0.96 [63] with support calculated from 100 bootstrap replicates. Bootstrap support (BS) values ≥ 70% were considered to have strong support. Partitioned Bayesian analyses (BA) were conducted using MrBayes v.3.1.2 [64], with two independent runs of Metropolis-coupled Markov chain Monte Carlo (MCMCMC) analyses, each with 4 chains and 1 million generations, with trees sampled every 100 generations. The level of convergence was assessed by checking the average standard deviation of split frequencies (<0.005). Convergence of the runs was also checked visually in Tracer ver. 1.5 [65], ensuring the effective sample sizes (ESS) were all above 200. Bayesian posterior probabilities (PP) were calculated by generating a 50% majority-rule consensus tree from the remaining sampled trees after discarding the burn-in (10%). PP values ≥ 0.95 indicate statistical
support. selleck products Detection of recombination and natural selection A codon-based approach implemented in HYPHY 2.0 [41] was used to analyze selection pressures within the seven individual protein-encoding genes, using a neighbor-joining model. Genetic algorithm recombination detection (GARD) was first used to identify any possible recombination breakpoints within each gene. Single likelihood ancestor counting (SLAC) was employed to calculate the global nonsynonymous (d N) and synonymous (d S) nucleotide substitution rate ratios (ω = d N/d S), with 95% confidence intervals; and to test the selection of variable codon sites based on the most appropriate nucleotide substitution model and tree topology, with a critical p-value of 0.05.