LoRaD: Marginal likelihood estimation with haste (but no waste).

The Lowest Radial Distance (LoRaD) method is a modification of the recently introduced Partition-Weighted Kernel method for estimating the marginal likelihood of a model, a quantity important for Bayesian model selection. For analyses involving a fixed tree topology, LoRaD improves upon the Steppingstone or Thermodynamic Integration (Path Sampling) approaches now in common use in phylogenetics because it requires sampling only from the posterior distribution, avoiding the need to sample from a series of ad hoc power posterior distributions, and yet is more accurate than other fast methods such as the Generalized Harmonic Mean (GHM) method. We show that the method performs well in comparison to the Generalized Steppingstone method on an empirical fixed-topology example from molecular phylogenetics involving 180 parameters. The LoRaD method can also be used to obtain the marginal likelihood in the variable-topology case if at least one tree topology occurs with sufficient frequency in the posterior sample to allow accurate estimation of the marginal likelihood conditional on that topology. [Bayesian; marginal likelihood; phylogenetics.].

An experimental test of Lanchester's models of combat in the neotropical termite Nasutitermes corniger (Blattodea: Termitidae).

Lanchester's models of combat have been invoked to explain the mechanics of group fighting in social animals. Specifically, Lanchester's square law posits that the fighting ability of the group is proportional to the square of the number of combatants. Although used to explain a variety of ecological phenomena, the models have not been thoroughly tested. We tested the Lanchester models using group battles between colonies of the termite Nasutitermes corniger. Our main goals were to determine if mortality rates fit the Lanchester models, and if so, whether the behavioural mechanisms underlying a group's success match those used in deriving the model. We initiated battles between pairs of colonies with different ratios of fighters and recorded deaths over time. We found that the numerically larger army has an advantage, but that the advantage is not as pronounced as predicted by Lanchester's square law. We also video-recorded battles to analyse individual behaviour, which did not support the mechanisms invoked by Lanchester. Instead, the killing power of an individual is increased by the presence of nest-mates, giving the larger group a disproportionate advantage. Although the behavioural mechanisms leading to the advantage may differ, our results still support some of the proposed ecological phenomena.

Dysphagia after Stroke: An Unmet Antibiotic Stewardship Opportunity.

The goal of antibiotic stewardship is to improve antibiotic use, often by reducing unnecessary treatment. Bedside dysphagia screening tools help identify patients at high risk of aspiration following stroke. Presence of dysphagia does not indicate a need for antibiotic treatment. Therefore, this retrospective, cohort study was developed to evaluate the association of dysphagia and antibiotic prescribing following stroke. There were 117 patients included. Patients were placed into 2 cohorts based on the results of the dysphagia screening, with 55 patients positive for dysphagia and 62 patients negative for dysphagia. Patients with dysphagia tended to be older, had higher National Institutes of Health stroke scores, and lower renal function. Patients with dysphagia were prescribed more empiric antibiotics than those without dysphagia (18.2% vs. 3.2%, p = 0.01). This resulted in 53 antibiotic days of therapy in the dysphagia cohort compared to 19 antibiotic days of therapy in the no dysphagia cohort (p = 0.1). No patients later developed pneumonia and only one patient was started antibiotics after 48 h. Two cases of Clostridioides difficile were reported. Both patients were in the dysphagia cohort and received antibiotics. Multivariable logistic regression demonstrated that positive chest x-ray findings and failed dysphagia screen were independent conditions associated with initiating antibiotics. These findings indicate that antibiotic use was higher in patients following stroke with a positive dysphagia screen. Close monitoring of stroke patients, particularly when positive for dysphagia, might be an under-recognized antibiotic stewardship opportunity.

Risk Factor Evaluation for Methicillin-Resistant Staphylococcus aureus and Pseudomonas aeruginosa in Community-Acquired Pneumonia.

BACKGROUND: The 2019 community-acquired pneumonia guidelines recommend using recent respiratory cultures and locally validated epidemiology plus risk factor assessment to determine empirical coverage of methicillin-resistant Staphylococcus aureus (MRSA) and Pseudomonas aeruginosa. OBJECTIVE: To develop a methodology for evaluating local epidemiology and validating local risk factors for P aeruginosa and MRSA. METHODS: This multicenter, retrospective cohort evaluated adult patients admitted for pneumonia. Risk factors for MRSA and P aeruginosa were evaluated using multivariable logistic regression and reported as adjusted odds ratios (aORs). RESULTS: There were 10 723 cases evaluated. Lung abscess/empyema had the highest odds associated with MRSA (aOR = 4.24; P < 0.0001), followed by influenza (aOR = 2.34; P = 0.01), end-stage renal disease (ESRD; aOR = 2.09; P = 0.006), illicit substance use (aOR = 1.7; P = 0.007), and chronic obstructive pulmonary disease (COPD; aOR = 1.26; P = 0.04). For P aeruginosa, the highest odds were in bronchiectasis (aOR = 6.13; P < 0.0001), lung abscess/empyema (aOR = 3.36; P = 0.005), and COPD (aOR = 1.84; P < 0.0001). Isolated COPD without other risk factors did not pose an increased risk of either organism. CONCLUSION AND RELEVANCE: Influenza, ESRD, lung abscess/empyema, and illicit substance use were local risk factors for MRSA. Bronchiectasis and lung abscess/empyema were risk factors for Pseudomonas. COPD was associated with MRSA and Pseudomonas. However, isolated COPD had similar rates of MRSA and Pseudomonas pneumonia compared with the total population. This study established a feasible methodology for evaluating local risk factors.

Utility of the respiratory viral panel as an antimicrobial stewardship tool.

WHAT IS KNOWN AND OBJECTIVE: The development of rapid diagnostics has revolutionized antimicrobial stewardship with efforts targeting earlier de-escalation or discontinuation of antibiotics. The respiratory viral panel (RVP) is one tool quickly able to detect common viral and bacterial pathogens using polymerase chain reaction technology. Utility may be further enhanced in conjunction with procalcitonin (PCT). However, the optimal use of the RVP to the clinical pharmacist in the treatment of community-acquired respiratory infections remains unclear. METHODS: The purpose of this guide is to review the available literature regarding the impact of the RVP with and without procalcitonin on antimicrobial stewardship efforts and to provide guidance on how to use each of these tools. RESULTS AND DISCUSSION: In total, 13 studies were included, 5 of which utilized PCT in conjunction with RVP and 8 of which did not use PCT. The majority of studies were retrospective in nature, and the most common outcomes evaluated were antibiotic days of therapy (DOT) and time to antibiotic discontinuation. WHAT IS NEW AND CONCLUSION: After review, RVP alone has limited value to antimicrobial stewardship; however, when used in conjunction with procalcitonin, RVP has the potential to reduce antibiotic use and duration.

Assessing Combinability of Phylogenomic Data Using Bayes Factors.

With the rapid reduction in sequencing costs of high-throughput genomic data, it has become commonplace to use hundreds of genes to infer phylogeny of any study system. While sampling a large number of genes has given us a tremendous opportunity to uncover previously unknown relationships and improve phylogenetic resolution, it also presents us with new challenges when the phylogenetic signal is confused by differences in the evolutionary histories of sampled genes. Given the incorporation of accurate marginal likelihood estimation methods into popular Bayesian software programs, it is natural to consider using the Bayes Factor (BF) to compare different partition models in which genes within any given partition subset share both tree topology and edge lengths. We explore using marginal likelihood to assess data subset combinability when data subsets have varying levels of phylogenetic discordance due to deep coalescence events among genes (simulated within a species tree), and compare the results with our recently described phylogenetic informational dissonance index (D) estimated for each data set. BF effectively detects phylogenetic incongruence and provides a way to assess the statistical significance of D values. We use BFs to assess data combinability using an empirical data set comprising 56 plastid genes from the green algal order Volvocales. We also discuss the potential need for calibrating BFs and demonstrate that BFs used in this study are correctly calibrated.

BEAGLE 3: Improved Performance, Scaling, and Usability for a High-Performance Computing Library for Statistical Phylogenetics.

BEAGLE is a high-performance likelihood-calculation library for phylogenetic inference. The BEAGLE library defines a simple, but flexible, application programming interface (API), and includes a collection of efficient implementations for calculation under a variety of evolutionary models on different hardware devices. The library has been integrated into recent versions of popular phylogenetics software packages including BEAST and MrBayes and has been widely used across a diverse range of evolutionary studies. Here, we present BEAGLE 3 with new parallel implementations, increased performance for challenging data sets, improved scalability, and better usability. We have added new OpenCL and central processing unit-threaded implementations to the library, allowing the effective utilization of a wider range of modern hardware. Further, we have extended the API and library to support concurrent computation of independent partial likelihood arrays, for increased performance of nucleotide-model analyses with greater flexibility of data partitioning. For better scalability and usability, we have improved how phylogenetic software packages use BEAGLE in multi-GPU (graphics processing unit) and cluster environments, and introduced an automated method to select the fastest device given the data set, evolutionary model, and hardware. For application developers who wish to integrate the library, we also have developed an online tutorial. To evaluate the effect of the improvements, we ran a variety of benchmarks on state-of-the-art hardware. For a partitioned exemplar analysis, we observe run-time performance improvements as high as 5.9-fold over our previous GPU implementation. BEAGLE 3 is free, open-source software licensed under the Lesser GPL and available at https://beagle-dev.github.io.

False negative diagnostic errors with polymerase chain reaction for the detection of cryptococcal meningoencephalitis.

The accuracy of the BioFire FilmArray Meningitis/Encephalitis (ME) panel for the identification of Cryptococcus has recently been called into question. The primary objective of this study was to assess the agreement between the BioFire ME polymerase chain reaction (PCR) and other markers of cryptococcal infection. This retrospective review identified five patients with cryptococcal meningoencephalitis, 4 of whom had a negative ME panel for Cryptococcus. All five cases had positive serum cryptococcal antigens, and three of five had a positive cerebrospinal fluid (CSF) culture for Cryptococcus. The BioFire ME panel does not appear to be reliable for ruling out Cryptococcus meningoencephalitis; multiple testing methods are recommended.

Peripheral neuropathy in non-Hodgkin's lymphoma patients receiving vincristine with and without aprepitant/fosaprepitant.

BACKGROUND: Peripheral neuropathy is a common treatment-related adverse effect associated with vincristine. Vincristine is a major CYP3A4 substrate and is often administered alongside the neurokinin-1 (NK-1) receptor antagonists, aprepitant or fosaprepitant, which are moderate CYP3A4 inhibitors. This inhibition may result in increased concentrations of vincristine and an increased incidence of toxicity. OBJECTIVE: The primary objective of this study was to investigate if there is a clinically significant drug interaction between vincristine and aprepitant or fosaprepitant resulting in early-onset peripheral neuropathy. The secondary objective of this study was to investigate the cumulative rate of chemotherapy-induced peripheral neuropathy (CIPN). METHODOLOGY: This was a single-centered, retrospective, cohort chart review. Patients receiving vincristine-based chemotherapy between 1 July 2010 through 30 June 2018 were identified and reviewed for concomitant use of aprepitant or fosaprepitant and incidence of neuropathy. Early-onset CIPN was defined as neuropathy onset during the first cycle of chemotherapy. RESULTS: A total of 115 subjects were retrospectively reviewed over the study period, of whom 71 were included in the aprepitant/fosaprepitant group and 44 were included in the group without a NK-1 receptor antagonist. Of the subjects who received aprepitant/fosaprepitant, 26.7% experienced early-onset peripheral neuropathy as compared to 22.7% in the group without a NK-1 receptor antagonist (P = 0.627). Overall, CIPN was higher in the group who received aprepitant/fosaprepitant compared to the group without (56% vs. 36%, P = 0.036). CONCLUSION: There appears to be an increased risk of CIPN with the concomitant use of vincristine and aprepitant or fosaprepitant.

Non-occlusive ST-segment elevated myocardial infarction following the administration of liposomal amphotericin B in the treatment of cryptococcal meningitis.

WHAT IS KNOWN AND OBJECTIVE: Liposomal amphotericin B (L-AmB) is the cornerstone of many serious invasive fungal infections. Despite lower frequencies of commonly reported adverse events in clinical trials compared to conventional formulations, post-marketing complications continue to mount. CASE DESCRIPTION: We present a case of chest pain following the initial dose of L-AmB for cryptococcal meningitis. Electrocardiogram demonstrated no acute electrocardiogram findings. Upon rechallenge, the chest pain worsened was subsequently accompanied by ST-segment elevation. Emergent coronary angiography found no acute findings. WHAT IS NEW AND CONCLUSION: Providers should be aware of cardiac complications with L-AmB, including non-occlusive ST-segment elevation.

Organellar phylogenomics inform systematics in the green algal family Hydrodictyaceae (Chlorophyceae) and provide clues to the complex evolutionary history of plastid genomes in the green algal tree of life.

PREMISE OF THE STUDY: Phylogenomic analyses across the green algae are resolving relationships at the class, order, and family levels and highlighting dynamic patterns of evolution in organellar genomes. Here we present a within-family phylogenomic study to resolve genera and species relationships in the family Hydrodictyaceae (Chlorophyceae), for which poor resolution in previous phylogenetic studies, along with divergent morphological traits, have precluded taxonomic revisions. METHODS: Complete plastome sequences and mitochondrial protein-coding gene sequences were acquired from representatives of the Hydrodictyaceae using next-generation sequencing methods. Plastomes were characterized, and gene order and content were compared with plastomes spanning the Sphaeropleales. Single-gene and concatenated-gene phylogenetic analyses of plastid and mitochondrial genes were performed. KEY RESULTS: The Hydrodictyaceae contain the largest sphaeroplealean plastomes thus far fully sequenced. Conservation of plastome gene order within Hydrodictyaceae is striking compared with more dynamic patterns revealed across Sphaeropleales. Phylogenetic analyses resolve Hydrodictyon sister to a monophyletic Pediastrum, though the morphologically distinct P. angulosum and P. duplex continue to be polyphyletic. Analyses of plastid data supported the neochloridacean genus Chlorotetraëdron as sister to Hydrodictyaceae, while conflicting signal was found in the mitochondrial data. CONCLUSIONS: A phylogenomic approach resolved within-family relationships not obtainable with previous phylogenetic analyses. Denser taxon sampling across Sphaeropleales is necessary to capture patterns in plastome evolution, and further taxa and studies are needed to fully resolve the sister lineage to Hydrodictyaceae and polyphyly of Pediastrum angulosum and P. duplex.

Treatment strategies for persistent methicillin-resistant Staphylococcus aureus bacteraemia.

WHAT IS KNOWN AND OBJECTIVE: Treatment of methicillin-resistant Staphylococcus aureus (MRSA) bacteraemia is a long-standing challenge to health care, often complicated by metastatic infections, treatment failure and mortality. When MRSA bacteraemia persists despite adequate initial treatment, current Infectious Diseases Society of America guidelines recommend evaluation and removal of possible sources of infection. In addition, a change in therapy may be considered. The objective of this review was to explore the therapeutic options for the treatment of persistent MRSA bacteraemia. METHODS: A literature search of PubMed, MEDLINE and Google Scholar was performed using the following search terms: [methicillin-resistant Staphylococcus aureus OR MRSA] AND [bacteraemia OR bloodstream infection] AND [persistent OR persistence OR refractory OR treatment failure OR salvage] AND treatment. We evaluated relevant, adult, English-language, peer-reviewed studies published between 1985 and May 2018. In vitro and animal studies were considered as supportive of in vivo data. RESULTS AND DISCUSSION: Randomized, controlled trials are lacking. However, case series and case reports support multiple treatment options including high-dose daptomycin in combination with an antistaphylococcal ß-lactam, ceftaroline, trimethoprim-sulfamethoxazole (TMP-SMX) or fosfomycin; ceftaroline alone or in combination with vancomycin or TMP-SMX; linezolid alone or in combination with a carbapenem, or telavancin. WHAT IS NEW AND CONCLUSION: Given the heterogeneity of the data, a preferred regimen has not emerged. Prescribers must take into consideration recent exposure, source control, and available synergy and clinical data. Further comparative trials are needed to establish a preferred regimen and the creation of a universal treatment algorithm.

Estimating Bayesian Phylogenetic Information Content.

Measuring the phylogenetic information content of data has a long history in systematics. Here we explore a Bayesian approach to information content estimation. The entropy of the posterior distribution compared with the entropy of the prior distribution provides a natural way to measure information content. If the data have no information relevant to ranking tree topologies beyond the information supplied by the prior, the posterior and prior will be identical. Information in data discourages consideration of some hypotheses allowed by the prior, resulting in a posterior distribution that is more concentrated (has lower entropy) than the prior. We focus on measuring information about tree topology using marginal posterior distributions of tree topologies. We show that both the accuracy and the computational efficiency of topological information content estimation improve with use of the conditional clade distribution, which also allows topological information content to be partitioned by clade. We explore two important applications of our method: providing a compelling definition of saturation and detecting conflict among data partitions that can negatively affect analyses of concatenated data. [Bayesian; concatenation; conditional clade distribution; entropy; information; phylogenetics; saturation.].

Anchored phylogenomics improves the resolution of evolutionary relationships in the rapid radiation of Protea L.

PREMISE OF THE STUDY: Estimating phylogenetic relationships in relatively recent evolutionary radiations is challenging, especially if short branches associated with recent divergence result in multiple gene tree histories. We combine anchored enrichment next-generation sequencing with species tree analyses to produce a robust estimate of phylogenetic relationships in the genus Protea (Proteaceae), an iconic radiation in South Africa. METHODS: We sampled multiple individuals within 59 out of 112 species of Protea and 6 outgroup species for a total of 163 individuals, and obtained sequences for 498 low-copy, orthologous nuclear loci using anchored phylogenomics. We compare several approaches for building species trees, and explore gene tree-species tree discrepancies to determine whether poor phylogenetic resolution reflects a lack of informative sites, incomplete lineage sorting, or hybridization. KEY RESULTS: Phylogenetic estimates from species tree approaches are similar to one another and recover previously well-supported clades within Protea, in addition to providing well-supported phylogenetic hypotheses for previously poorly resolved intrageneric relationships. Individual gene trees are markedly different from one another and from species trees. Nonetheless, analyses indicate that differences among gene trees occur primarily concerning clades supported by short branches. CONCLUSIONS: Species tree methods using hundreds of nuclear loci provided strong support for many previously unresolved relationships in the radiation of the genus Protea. In cases where support for particular relationships remains low, these appear to arise from few informative sites and lack of information rather than strongly supported disagreement among gene trees.

Evolution of woody life form on tropical mountains in the tribe Spermacoceae (Rubiaceae).

PREMISE OF THE STUDY: Spermacoceae are mainly an herbaceous group in the Rubiaceae. However, a few lineages are woody and are found in a diverse range of habitat types. Three of the largest woody lineages (Arcytophyllum, Hedyotis, and Kadua) are characterized by their distribution in the moist tropical mountains and have disjunct distribution patterns with respect to their closest relatives. In this study, we explore the cases of derived woodiness in these three lineages and their diversification dynamics in the tropical mountains of Asia, the Pacific, and the Americas. METHODS: By combining phylogenetic results with wood anatomical studies, we estimated timing of origin of the three woody groups, inferred their ancestral traits and ancestral distribution ranges, analyzed their associations with the tropical upland habitat, and elucidated their diversification across tropical mountains. KEY RESULTS: The three woody clades originated and diversified from herbaceous ancestors in close association with the tropical upland habitat during the Miocene. The ancestral range for Asian-Pacific Hedyotis is Africa/Madagascar and continental Asia for Pacific Kadua. The complex geological history of tropical Asia allowed Hedyotis to diversify faster and create narrow endemics near oceans in the highlands of the Western Ghats in India, Sri Lanka, Southeast Asia including southeastern China, and New Guinea. CONCLUSIONS: The three major woody clades in Spermacoceae have gained their woodiness independently from one another, subsequent to colonization by their ancestors from a different geographic environment. The evolution and diversification along the tropical mountain orogeny is strongly linked with the formation of woody habit and many narrow endemic species.

Implementation of Global Strategies to Prevent Hospital-Onset Clostridium difficile Infection: Targeting Proton Pump Inhibitors and Probiotics.

BACKGROUND: Proton pump inhibitors (PPIs) have been identified as a significant risk factor for the development of Clostridium difficile infection (CDI). Probiotics given concurrently with antibiotics have been shown to have a moderate impact on preventing CDI. OBJECTIVE: To evaluate the effectiveness of hospital-wide interventions designed to reduce PPI use and increase probiotics and whether these interventions were associated with a change in the incidence of hospital onset (HO)-CDI. METHODS: This retrospective cohort study compared 2 fiscal years: July 2013 to June 2014 (FY14) and July 2014 to June 2015 (FY15). In July of FY15, global educational initiatives were launched targeting PPIs. Additionally, a HO-CDI prevention bundle was added to antibiotic-containing order sets targeting probiotics. Overall PPI use, probiotic use, and incidence of HO-CDI were recorded and compared for each cohort. Charts were also reviewed for patients who developed HO-CDI for the presence and appropriateness of a PPI and presence of probiotics. RESULTS: The interventions resulted in a decrease in PPI use by 14% or 96 doses/1000 patient days (TPD; P = 0.0002) and a reduction in IV PPI use by 31% or 71 doses/TPD ( P = 0.0008). Probiotic use increased by 130% or 126 doses/TPD ( P = 0.0006). The incidence of HO-CDI decreased by 20% or 0.1 cases/TPD ( P = 0.04). CONCLUSIONS: A collaborative, multifaceted educational initiative directed at highlighting the risks associated with PPI use was effective in reducing PPI prescribing. The implementation of a probiotic bundle added to antibiotic order sets was effective in increasing probiotic use. These interventions were associated with a decrease in incidence of HO-CDI.

Chloroplast phylogenomic data from the green algal order Sphaeropleales (Chlorophyceae, Chlorophyta) reveal complex patterns of sequence evolution.

Chloroplast sequence data are widely used to infer phylogenies of plants and algae. With the increasing availability of complete chloroplast genome sequences, the opportunity arises to resolve ancient divergences that were heretofore problematic. On the flip side, properly analyzing large multi-gene data sets can be a major challenge, as these data may be riddled with systematic biases and conflicting signals. Our study contributes new data from nine complete and four fragmentary chloroplast genome sequences across the green algal order Sphaeropleales. Our phylogenetic analyses of a 56-gene data set show that analyzing these data on a nucleotide level yields a well-supported phylogeny - yet one that is quite different from a corresponding amino acid analysis. We offer some possible explanations for this conflict through a range of analyses of modified data sets. In addition, we characterize the newly sequenced genomes in terms of their structure and content, thereby further contributing to the knowledge of chloroplast genome evolution.

Phycas: software for Bayesian phylogenetic analysis.

Phycas is open source, freely available Bayesian phylogenetics software written primarily in C++ but with a Python interface. Phycas specializes in Bayesian model selection for nucleotide sequence data, particularly the estimation of marginal likelihoods, central to computing Bayes Factors. Marginal likelihoods can be estimated using newer methods (Thermodynamic Integration and Generalized Steppingstone) that are more accurate than the widely used Harmonic Mean estimator. In addition, Phycas supports two posterior predictive approaches to model selection: Gelfand-Ghosh and Conditional Predictive Ordinates. The General Time Reversible family of substitution models, as well as a codon model, are available, and data can be partitioned with all parameters unlinked except tree topology and edge lengths. Phycas provides for analyses in which the prior on tree topologies allows polytomous trees as well as fully resolved trees, and provides for several choices for edge length priors, including a hierarchical model as well as the recently described compound Dirichlet prior, which helps avoid overly informative induced priors on tree length.

Hiding in plain sight: Koshicola spirodelophila gen. et sp. nov. (Chaetopeltidales, Chlorophyceae), a novel green alga associated with the aquatic angiosperm Spirodela polyrhiza.

PREMISE OF THE STUDY: Discovery and morphological characterization of a novel epiphytic aquatic green alga increases our understanding of Chaetopeltidales, a poorly known order in Chlorophyceae. Chloroplast genomic data from this taxon reveals an unusual architecture previously unknown in green algae. METHODS: Using light and electron microscopy, we characterized the morphology and ultrastructure of a novel taxon of green algae. Bayesian phylogenetic analyses of nuclear and plastid genes were used to test the hypothesized membership of this taxon in order Chaetopeltidales. With next-generation sequence data, we assembled the plastid genome of this novel taxon and compared its gene content and architecture to that of related species to further investigate plastid genome traits. KEY RESULTS: The morphology and ultrastructure of this alga are consistent with placement in Chaetopeltidales (Chlorophyceae), but a distinct trait combination supports recognition of this alga as a new genus and species-Koshicola spirodelophila gen. et sp. nov. Its placement in the phylogeny as a descendant of a deep division in the Chaetopeltidales is supported by analysis of molecular data sets. The chloroplast genome is among the largest reported in green algae and the genes are distributed on three large (rather than a single) chromosome, in contrast to other studied green algae. CONCLUSIONS: The discovery of Koshicola spirodelophila gen. et sp. nov. highlights the importance of investigating even commonplace habitats to explore new microalgal diversity. This work expands our understanding of the morphological and chloroplast genomic features of green algae, and in particular those of the poorly studied Chaetopeltidales.

Posterior predictive Bayesian phylogenetic model selection.

We present two distinctly different posterior predictive approaches to Bayesian phylogenetic model selection and illustrate these methods using examples from green algal protein-coding cpDNA sequences and flowering plant rDNA sequences. The Gelfand-Ghosh (GG) approach allows dissection of an overall measure of model fit into components due to posterior predictive variance (GGp) and goodness-of-fit (GGg), which distinguishes this method from the posterior predictive P-value approach. The conditional predictive ordinate (CPO) method provides a site-specific measure of model fit useful for exploratory analyses and can be combined over sites yielding the log pseudomarginal likelihood (LPML) which is useful as an overall measure of model fit. CPO provides a useful cross-validation approach that is computationally efficient, requiring only a sample from the posterior distribution (no additional simulation is required). Both GG and CPO add new perspectives to Bayesian phylogenetic model selection based on the predictive abilities of models and complement the perspective provided by the marginal likelihood (including Bayes Factor comparisons) based solely on the fit of competing models to observed data.