Citation Context

...ed employing the same model used by R2010 (nt-GTR without codon partitioning). A Bayesian analysis of the nt data set was also performed using the nt-CAT model, and 10-fold Bayesian cross-validation (=-=Stone 1974-=-) was used to test whether nt-GTR or nt-CAT fits this data set better. For the cross-validation, we used a training set composed of 90% of the sites in the alignment and a test set composed of the rem...

Citation Context

... al. 2009). For both programs, two independent runs were performed and, in the caseofMrBayes, each runused4differentially heated chains.All BayesianGTR (Lanave et al. 1984; Yang et al. 1998) and JTT (=-=Jones et al. 1992-=-) analyses were performedusingMrBayes.All analyses performedusing the CAT and the CATGTR model (Lartillot and Philippe 2004) were performed using Phylobayes. Analyses were stoppedwhen the standarddevi...

Citation Context

...ekar et al. 2004) and PhyloBayes3 (Lartillot et al. 2009). For both programs, two independent runs were performed and, in the caseofMrBayes, each runused4differentially heated chains.All BayesianGTR (=-=Lanave et al. 1984-=-; Yang et al. 1998) and JTT (Jones et al. 1992) analyses were performedusingMrBayes.All analyses performedusing the CAT and the CATGTR model (Lartillot and Philippe 2004) were performed using Phylobay...

Citation Context

... the fast ML method to obtain bootstrap support (100 replicates). Bayesian analyseswere conductedusing theMPI version of MrBayes3 (Ronquist and Huelsenbeck 2003; Altekar et al. 2004) and PhyloBayes3 (=-=Lartillot et al. 2009-=-). For both programs, two independent runs were performed and, in the caseofMrBayes, each runused4differentially heated chains.All BayesianGTR (Lanave et al. 1984; Yang et al. 1998) and JTT (Jones et ...

Citation Context

...d PhyloBayes3 (Lartillot et al. 2009). For both programs, two independent runs were performed and, in the caseofMrBayes, each runused4differentially heated chains.All BayesianGTR (Lanave et al. 1984; =-=Yang et al. 1998-=-) and JTT (Jones et al. 1992) analyses were performedusingMrBayes.All analyses performedusing the CAT and the CATGTR model (Lartillot and Philippe 2004) were performed using Phylobayes. Analyses were ...

Citation Context

...tional analyses.— Codon usage has been estimated using GCUA (McInerney 1998). We quantified taxon-specific biases in synonymous, serine codon families (TCNorAGY)usage, with a statistic based on skew (=-=Perna and Kocher 1995-=-). This statistic, named TCN/AGY skew, is calculated as (TCN -AGY)/(TCN+AGY) and ranges from +1 to −1. If TCN/AGY = 1 only TCN codons are used. If TCN/AGY = −1 only AGY codons are used. If TCN/AGY = 0...

Citation Context

...(Foster et al. 1997; Saccone et al. 1999). In such circumstances, the time-homogeneity assumption of stationary models is violated (Lockhart et al. 1992; Galtier and Gouy 1995; Yang and Roberts 1995; =-=Foster 2004-=-; Jermiin et al. 2004; Gibson et al. 2005; Blanquart and Lartillot 2008; Foster et al. 2009), and unless time-heterogeneous models are employed, compositional attractions [see Jeffroy et al. (2006) fo...

Citation Context

...of the Signal in the nt and aa Data Sets Use of the slow–fast method.—To explore the nature of the signal in the nt and aa data sets, we excluded fast evolving sites using a slow–fast-based approach (=-=Brinkmann and Phillipe 1999-=-). The evolutionary rate of each site in the aa and nt alignments was estimated as its cumulative Parsimony score (P-score). This is obtained by summing, for each character, the P-scores calculated on...

Citation Context

... in unrelated lineages (Foster et al. 1997; Saccone et al. 1999). In such circumstances, the time-homogeneity assumption of stationary models is violated (Lockhart et al. 1992; Galtier and Gouy 1995; =-=Yang and Roberts 1995-=-; Foster 2004; Jermiin et al. 2004; Gibson et al. 2005; Blanquart and Lartillot 2008; Foster et al. 2009), and unless time-heterogeneous models are employed, compositional attractions [see Jeffroy et ...

Citation Context

...of specific nucleotides in unrelated lineages (Foster et al. 1997; Saccone et al. 1999). In such circumstances, the time-homogeneity assumption of stationary models is violated (Lockhart et al. 1992; =-=Galtier and Gouy 1995-=-; Yang and Roberts 1995; Foster 2004; Jermiin et al. 2004; Gibson et al. 2005; Blanquart and Lartillot 2008; Foster et al. 2009), and unless time-heterogeneous models are employed, compositional attra...

Citation Context

...recoding strategies (e.g., R-Y coding as in Woese et al. 1991). Because saturation and mutational pressures preferentially affect synonymous sites, amino acid data sets are expected to be less prone (=-=Jeffroy et al. 2006-=-; Rota-Stabelli et al. 2010), but not immune (Foster and Hickey 1999), to both problems. Codon-usage biases can also cause phylogenetic errors in nucleotide-based data sets. (Inagaki and Roger, 2006) ...

Citation Context

...ecause saturation and mutational pressures preferentially affect synonymous sites, amino acid data sets are expected to be less prone (Jeffroy et al. 2006; Rota-Stabelli et al. 2010), but not immune (=-=Foster and Hickey 1999-=-), to both problems. Codon-usage biases can also cause phylogenetic errors in nucleotide-based data sets. (Inagaki and Roger, 2006) and Inagaki et al. (2004) showed that, in the case of deep divergenc...

Citation Context

...cumstances, the time-homogeneity assumption of stationary models is violated (Lockhart et al. 1992; Galtier and Gouy 1995; Yang and Roberts 1995; Foster 2004; Jermiin et al. 2004; Gibson et al. 2005; =-=Blanquart and Lartillot 2008-=-; Foster et al. 2009), and unless time-heterogeneous models are employed, compositional attractions [see Jeffroy et al. (2006) for review] can sway the results of phylogenetic analyses. Common practic...

Citation Context

...sponding (i.e., translated) amino acid sequences, finding that many of the crustacean relationships supported by the nucleotides (see also Regier and Zwick 2011) are not found when using amino acids (=-=Regier et al. 2010-=-). Hence, R2010’s data set represents an extremely interesting case to evaluate the issue of data-type choice in phylogenomics. The R2010 nt analyses (under both codon and nucleotide models) identifie...

Citation Context

...ults of phylogenetic analyses. Common practice to overcome these problems is to remove third-codon positions from nucleotide data sets, or to use character recoding strategies (e.g., R-Y coding as in =-=Woese et al. 1991-=-). Because saturation and mutational pressures preferentially affect synonymous sites, amino acid data sets are expected to be less prone (Jeffroy et al. 2006; Rota-Stabelli et al. 2010), but not immu...

Citation Context

.... 1999). In such circumstances, the time-homogeneity assumption of stationary models is violated (Lockhart et al. 1992; Galtier and Gouy 1995; Yang and Roberts 1995; Foster 2004; Jermiin et al. 2004; =-=Gibson et al. 2005-=-; Blanquart and Lartillot 2008; Foster et al. 2009), and unless time-heterogeneous models are employed, compositional attractions [see Jeffroy et al. (2006) for review] can sway the results of phyloge...

Citation Context

...e biased accumulation of specific nucleotides in unrelated lineages (Foster et al. 1997; Saccone et al. 1999). In such circumstances, the time-homogeneity assumption of stationary models is violated (=-=Lockhart et al. 1992-=-; Galtier and Gouy 1995; Yang and Roberts 1995; Foster 2004; Jermiin et al. 2004; Gibson et al. 2005; Blanquart and Lartillot 2008; Foster et al. 2009), and unless time-heterogeneous models are employ...

Citation Context

...d. Alternatively, the use of non-stationary models (Foster 2004, Blanquart andLartillot 2008) shouldbe considered. In the long run, the use of codon models that can handle heterogeneity across sites (=-=Rodrigue et al. 2010-=-) should also be considered, but at this stage, the implementation of these models is still intractable. SUPPLEMENTARY MATERIAL Supplementary material, including data files and/or online-only appendic...

Citation Context

...under strong mutational pressure, and compositionally driven mutational pressure has previously been shown to be responsible for the biased accumulation of specific nucleotides in unrelated lineages (=-=Foster et al. 1997-=-; Saccone et al. 1999). In such circumstances, the time-homogeneity assumption of stationary models is violated (Lockhart et al. 1992; Galtier and Gouy 1995; Yang and Roberts 1995; Foster 2004; Jermii...

Citation Context

...are more often retained, mainly because of the implieddouble transversions, a putatively rare event. There are 2 ways of moving between the 2 TCN/AGY serine codon types: 2 simultaneous transversions (=-=Averof et al. 2000-=-) or 2 consecutive transversions through an intermediate non-serine state. Inspection of the 2204 positions in the alignment at which at least 2 taxa bear serine reveals that more than half of those p...

Citation Context

...udy Uncertainties remain about the internal relationships within the arthropod subphyla, particularly Pancrustacea (Crustacea plus Hexapoda). Crustacea are most likely paraphyletic (Cook et al. 2005; =-=Regier et al. 2005-=-), but the details of their relationships are highly debated. On the basis of respiratory proteins, Ertas et al. (2009) suggested that Remipedia—a small group of anchialine cave crustaceans—are the si...

Citation Context

...nd ESTs, which support Branchiopoda, and in some cases Copepoda, to be more closely related to Hexapoda than to Malacostraca (Mallatt and Giribet 2006; Von Reumont et al. 2009; Meusemann et al. 2010; =-=Campbell et al. 2011-=-; Rota-Stabelli et al. 2011). Here, we addressed the 2-fold problem of data-type choice and crustacean relationships, further exploring the signals in the R2010 data set. We performed a variety of ana...

Citation Context

...stance, to tRNA imbalance, mutational pressure, or a combination of both phenomena. In particular, mutational pressures are a well-known source of systematic errors (e.g., in bacterial phylogenetics; =-=Embley et al. 1993-=-; Cummins and McInerney 2011). We initially tested if the synonymous codon usage for serine (but also leucine and arginine) is homogeneously distributed among taxa. To measure synonymous codon usage, ...

Citation Context

...rone (Jeffroy et al. 2006; Rota-Stabelli et al. 2010), but not immune (Foster and Hickey 1999), to both problems. Codon-usage biases can also cause phylogenetic errors in nucleotide-based data sets. (=-=Inagaki and Roger, 2006-=-) and Inagaki et al. (2004) showed that, in the case of deep divergences among the eukaryotic lineages, phylogenetic analyses based on nucleotide sequences of plastid-encoded genes could be misled by ...

Citation Context

...rine-free. The data sets were analyzed using both nt-CAT (the best fitting model—see above) and nt-GTR. Codon-usage bias, skew, and compositional analyses.— Codon usage has been estimated using GCUA (=-=McInerney 1998-=-). We quantified taxon-specific biases in synonymous, serine codon families (TCNorAGY)usage, with a statistic based on skew (Perna and Kocher 1995). This statistic, named TCN/AGY skew, is calculated a...