Bayesian Program Research Articles

Differing lifestyles of Staphylococcus epidermidis as revealed through Bayesian clustering of multilocus sequence types

Staphylococcus epidermidis is part of the normal bacterial flora of human skin and a leading cause of infections associated with indwelling medical devices. Previous phylogenetic analyses of subgenomic data have been unable to distinguish between S. epidermidis strains with nosocomial or commensal lifestyles, despite the identification of specific phenotypes and accessory genes that may contribute to such lifestyles. To attempt to better define the population structure of this species, the international S. epidermidis multilocus sequence typing database was analyzed with the Bayesian clustering programs STRUCTURE and BAPS. A total of six genetic clusters (GCs) were identified. A local population of S. epidermidis from clinical specimens was classified according to these six GCs, and further characterized for antibiotic susceptibilities, biofilm, and various genetic markers. GC5 was abundant and significantly enriched for isolates that were resistant to four classes of antibiotics, high biofilm production, and positive for the virulence markers icaA, IS256, and sesD/bhp, indicating its potential clinical relevance. In contrast, GC2 was rare and contained the only isolates positive for the putative commensal marker, fdh. GC1 and GC6 were abundant but not significantly associated with any of the examined characteristics, except for sesF/aap and GC6. GC3 was rare and identified as a potential genetic sink that received, but did not donate, core genetic material from other GCs. In conclusion, population genetics analyses were essential for identifying clusters of strains that may differ in their adaptation to nosocomial or commensal lifestyles. These results provide a new, population genetics framework for studying S. epidermidis.

Dating Phylogenies with Sequentially Sampled Tips

We develop a Bayesian Markov chain Monte Carlo (MCMC) algorithm for estimating divergence times using sequentially sampled molecular sequences. This type of data is commonly collected during viral epidemics and is sometimes available from different species in ancient DNA studies. We derive the distribution of ages of nodes in the tree under a birth-death-sequential-sampling (BDSS) model and use it as the prior for divergence times in the dating analysis. We implement the prior in the MCMCtree program in the PAML package for divergence dating. The BDSS prior is very flexible and, with different parameters, can generate trees of very different shapes, suitable for examining the sensitivity of posterior time estimates. We apply the method to a data set of SIV/HIV-2 genes in comparison with a likelihood-based dating method, and to a data set of influenza H1 genes from different hosts in comparison with the Bayesian program BEAST. We examined the impact of tree topology on time estimates and suggest that multifurcating consensus trees should be avoided in dating analysis. We found posterior time estimates for old nodes to be sensitive to the priors on times and rates and suggest that previous Bayesian dating studies may have produced overconfident estimates.

Population Decline in a Long-Lived Species: The Wood Turtle in Michigan

Abstract: Populations of Wood Turtles, Glyptemys insculpta, have steadily decreased over the past 30 yr because of habitat destruction and degradation. We sampled Wood Turtles from three areas in Michigan, USA, to characterize populations, quantify demographic trends, and measure the effect of declining population size on genetic diversity. Wood Turtle samples (n = 68) were collected from three rivers in the Lower Peninsula of Michigan and analyzed at nine microsatellite loci. Bayesian clustering programs identified two populations that split the three sampling sites into North and South populations. In both populations, analysis of genealogies estimated r < 0, indicating population decline. However, no evidence of a bottleneck was detected (P = 0.30 North, P = 0.29 South), and little evidence of inbreeding was observed (average North FIS = 0.25, average South FIS = 0.23), relative to other Emydidae populations. The high genetic diversity observed in the North and South populations is likely due to immigr...

A Microbial Signature Approach to Identify Fecal Pollution in the Waters Off an Urbanized Coast of Lake Michigan

Urban coasts receive watershed drainage from ecosystems that include highly developed lands with sewer and stormwater infrastructure. In these complex ecosystems, coastal waters are often contaminated with fecal pollution, where multiple delivery mechanisms that often contain multiple fecal sources make it difficult to mitigate the pollution. Here, we exploit bacterial community sequencing of the V6 and V6V4 hypervariable regions of the bacterial 16S rRNA gene to identify bacterial distributions that signal the presence of sewer, fecal, and human fecal pollution. The sequences classified to three sewer infrastructure-associated bacterial genera, Acinetobacter, Arcobacter, and Trichococcus, and five fecal-associated bacterial families, Bacteroidaceae, Porphyromonadaceae, Clostridiaceae, Lachnospiraceae, and Ruminococcaceae, served as signatures of sewer and fecal contamination, respectively. The human fecal signature was determined with the Bayesian source estimation program SourceTracker, which we applied to a set of 40 sewage influent samples collected in Milwaukee, WI, USA to identify operational taxonomic units (≥ 97 % identity) that were most likely of human fecal origin. During periods of dry weather, the magnitudes of all three signatures were relatively low in Milwaukee's urban rivers and harbor and nearly zero in Lake Michigan. However, the relative contribution of the sewer and fecal signature frequently increased to > 2 % of the measured surface water communities following sewer overflows. Also during combined sewer overflows, the ratio of the human fecal pollution signature to the fecal pollution signature in surface waters was generally close to that of sewage, but this ratio decreased dramatically during dry weather and rain events, suggesting that nonhuman fecal pollution was the dominant source during these weather-driven scenarios. The qPCR detection of two human fecal indicators, human Bacteroides and Lachno2, confirmed the urban fecal footprint in this ecosystem extends to at least 8 km offshore.

PHC-020 Pharmacokinetically Guided Dose Adjustment of 5-Fluorouracil (5-FU) in Gastrointestinal Cancer Patients

BackgroundAppropriate dosing of chemotherapeutic drugs is critical to reducing mortality and increasing progression-free survival. 5-fluorouracil (5-FU) is a widely used chemotherapeutic drug in gastrointestinal cancer. The standard approach to dosing...

A hybrid dilemma: a molecular investigation of South African bontebok (Damaliscus pygargus pygargus) and blesbok (Damaliscus pygargus phillipsi)

Human-mediated hybridization poses a serious threat to the genetic integrity of threatened inter-fertile species or subspecies, such as the bontebok (Damaliscus pygargus pygargus) and blesbok (Damaliscus pygargus phillipsi) in South Africa. The identification of pure and admixed populations is key to sound biodiversity conservation management and practices. Where identification based on morphometric measurements are complex, molecular analysis can provide a tool to identify hybrids or introgression in populations. This study provides the first molecular analyses on pure bontebok and blesbok as well as putative hybrid populations and samples of unknown purity. Analysis of a panel of 13 microsatellite markers by applying a Bayesian model-based clustering programme provided evidence of clear partitioning between the two subspecies. Hybridization was detected in 33 % (40 of 121) of the samples with unknown purity. Currently, bontebok is listed as protected under the Convention on International Trade in Endangered Species of Wild Fauna and Flora and is listed as endangered under the United States of America’s Endangered Species Act. Our research found that the microsatellite markers used in this study could accurately identify pure and hybrid animals and can be used as a tool to conserve the genetic integrity of both subspecies.

Bayesian Approach to 14C Dates for Estimation of Long-Term Archaeological Sequences in Arid Environments: The Holocene Site of Takarkori Rockshelter, Southwest Libya

Caves and rockshelters are critical loci for the analysis and understanding of human trajectories in the past. Use and re-uses of the same context, however, might have had serious impacts on depositional aspects. This is particularly true for the archaeological history of desert environments, such as the central Sahara, where most of the deposits are made of loose sand, rich in organic matter. Besides traditional stratigraphic reconstructions and a detailed study of the material culture, radiocarbon measurements from different contexts analyzing several types of material (bone, dried and charred coprolite, uncharred and charred plant remains, etc.) can highlight intrinsic critical aspects of 14C determinations. These measurements must be carefully evaluated to provide a correct chronological assessment of the life history of the site. We present the statistics derived from the set of about 50 14C measurements from the site of Takarkori, southwest Libya, where early Holocene foragers and then groups of cattle herders inhabited the area from ∼10,200 to 4600 cal yr BP. We have used the BCal Bayesian 14C calibration program designed for statistical presentation of the calibrated data and the estimation of their probability for different phases. Results indicate that the Takarkori rockshelter was occupied during 4 phases of the following cultures: Late Acacus from 10,170 to 8180 cal yr BP; Early Pastoral, 8180–6890 cal yr BP; Middle Pastoral, 7160–5610 cal yr BP; and Late Pastoral, 5700–4650 cal yr BP.

The geographical and environmental determinants of genetic diversity for four alpine conifers of the European Alps

Climate is one of the most important drivers of local adaptation in forest tree species. Standing levels of genetic diversity and structure within and among natural populations of forest trees are determined by the interplay between climatic heterogeneity and the balance between selection and gene flow. To investigate this interplay, single nucleotide polymorphisms (SNPs) were genotyped in 24 to 37 populations from four subalpine conifers, Abies alba Mill., Larix decidua Mill., Pinus cembra L. and Pinus mugo Turra, across their natural ranges in the Italian Alps and Apennines. Patterns of population structure were apparent using a Bayesian clustering program, STRUCTURE, which identified three to five genetic groups per species. Geographical correlates with these patterns, however, were only apparent for P.cembra. Multivariate environmental variables [i.e. principal components (PCs)] were subsequently tested for association with SNPs using a Bayesian generalized linear mixed model. The majority of the SNPs, ranging from six in L.decidua to 18 in P.mugo, were associated with PC1, corresponding to winter precipitation and seasonal minimum temperature. In A.alba, four SNPs were associated with PC2, corresponding to the seasonal minimum temperature. Functional annotation of those genes with the orthologs in Arabidopsis revealed several genes involved in abiotic stress response. This study provides a detailed assessment of population structure and its association with environment and geography in four coniferous species in the Italian mountains.

Phylogenetics, likelihood, evolution and complexity

Phylogenetics, likelihood, evolution and complexity (PLEX) is a flexible and fast Bayesian Markov chain Monte Carlo software program for large-scale analysis of nucleotide and amino acid data using complex evolutionary models in a phylogenetic framework. The program gains large speed improvements over standard approaches by implementing 'partial sampling of substitution histories', a data augmentation approach that can reduce data analysis times from months to minutes on large comparative datasets. A variety of nucleotide and amino acid substitution models are currently implemented, including non-reversible and site-heterogeneous mixture models. Due to efficient algorithms that scale well with data size and model complexity, PLEX can be used to make inferences from hundreds to thousands of taxa in only minutes on a desktop computer. It also performs probabilistic ancestral sequence reconstruction. Future versions will support detection of co-evolutionary interactions between sites, probabilistic tests of convergent evolution and rigorous testing of evolutionary hypotheses in a Bayesian framework. PLEX v1.0 is licensed under GPL. Source code and documentation will be available for download at www.evolutionarygenomics.com/ProgramsData/PLEX. PLEX is implemented in C++ and supported on Linux, Mac OS X and other platforms supporting standard C++ compilers. Example data, control files, documentation and accessory Perl scripts are available from the website. David.Pollock@UCDenver.edu. Supplementary data are available at Bioinformatics online.

Bayesian models for comparative analysis integrating phylogenetic uncertainty

BackgroundUncertainty in comparative analyses can come from at least two sources: a) phylogenetic uncertainty in the tree topology or branch lengths, and b) uncertainty due to intraspecific variation in trait values, either due to measurement error or natural individual variation. Most phylogenetic comparative methods do not account for such uncertainties. Not accounting for these sources of uncertainty leads to false perceptions of precision (confidence intervals will be too narrow) and inflated significance in hypothesis testing (e.g. p-values will be too small). Although there is some application-specific software for fitting Bayesian models accounting for phylogenetic error, more general and flexible software is desirable.MethodsWe developed models to directly incorporate phylogenetic uncertainty into a range of analyses that biologists commonly perform, using a Bayesian framework and Markov Chain Monte Carlo analyses.ResultsWe demonstrate applications in linear regression, quantification of phylogenetic signal, and measurement error models. Phylogenetic uncertainty was incorporated by applying a prior distribution for the phylogeny, where this distribution consisted of the posterior tree sets from Bayesian phylogenetic tree estimation programs. The models were analysed using simulated data sets, and applied to a real data set on plant traits, from rainforest plant species in Northern Australia. Analyses were performed using the free and open source software OpenBUGS and JAGS.ConclusionsIncorporating phylogenetic uncertainty through an empirical prior distribution of trees leads to more precise estimation of regression model parameters than using a single consensus tree and enables a more realistic estimation of confidence intervals. In addition, models incorporating measurement errors and/or individual variation, in one or both variables, are easily formulated in the Bayesian framework. We show that BUGS is a useful, flexible general purpose tool for phylogenetic comparative analyses, particularly for modelling in the face of phylogenetic uncertainty and accounting for measurement error or individual variation in explanatory variables. Code for all models is provided in the BUGS model description language.

PHC023 Are the official recommendations for amikacine serum levels suitable for older patients?

BackgroundThe aminoglycosides represent the mainstay in the treatment of serious Gram-negative infections. Their use is difficult in older patients because of high potential toxicity and the large observed inter-individual variability....

Are the official recommendations for amikacine serum levels suitable for older patients?

BackgroundThe aminoglycosides represent the mainstay in the treatment of serious Gram-negative infections. Their use is difficult in older patients because of high potential toxicity and the large observed inter-individual variability....

Dating the appearance of Lapita pottery in the Bismarck Archipelago and its dispersal to Remote Oceania

ABSTRACT The Bayesian calibration program OxCal v.4.1.5 is applied to two chronological datasets for early Lapita derived from two comprehensive reviews. The two datasets are supplemented by published ages for early Lapita sites in two key island groups within Remote Oceania: Vanuatu and Fiji. The analyses provide statistically robust chronologies for the emergence of Lapita on Mussau at 3470–3250 cal BP and in the rest of the Bismarck Archipelago at 3360–3240 cal BP. After a period of 130–290 years, Lapita dispersed to Vanuatu by 3250–3100 cal BP and to Fiji by 3130–3010 cal BP.

Using PK/PD to Optimize Antibiotic Dosing for Critically Ill Patients

Antibiotic prescription for critically ill patients is a complicated process because of the pharmacokinetic differences of this patient population with non-critically ill patients and the lack of robust informative studies. This article seeks to review the available literature describing dosing requirements for optimized treatment of critically ill patients and to discuss a framework to rationally address complex cases by outlining the suggested processes for optimal achievement of pharmacokinetic/ pharmacodynamic targets. A variety of papers exist describing the effect of pathophysiology on antibiotic kinetics. In the critically ill patient, dysfunction of almost any organ system can result in significant changes to drug volume of distribution and clearance. Dysfunction of the cardiovascular and renal systems in particular is problematic and can lead to potentially sub-therapeutic antibiotic concentrations in blood and in interstitial fluid. In response to altered pharmacokinetics, dose regimens that adhere to the pharmacodynamics of the antibiotic are essential. In the absence of validated dosing algorithms, therapeutic drug monitoring data and susceptibility data of the infecting pathogen should be inputted into a Bayesian software program that include population pharmacokinetic models to calculate dosing regimens that are personalized for the critically ill patient.

Tail Paradox, Partial Identifiability, and Influential Priors in Bayesian Branch Length Inference

Recent studies have observed that Bayesian analyses of sequence data sets using the program MrBayes sometimes generate extremely large branch lengths, with posterior credibility intervals for the tree length (sum of branch lengths) excluding the maximum likelihood estimates. Suggested explanations for this phenomenon include the existence of multiple local peaks in the posterior, lack of convergence of the chain in the tail of the posterior, mixing problems, and misspecified priors on branch lengths. Here, we analyze the behavior of Bayesian Markov chain Monte Carlo algorithms when the chain is in the tail of the posterior distribution and note that all these phenomena can occur. In Bayesian phylogenetics, the likelihood function approaches a constant instead of zero when the branch lengths increase to infinity. The flat tail of the likelihood can cause poor mixing and undue influence of the prior. We suggest that the main cause of the extreme branch length estimates produced in many Bayesian analyses is the poor choice of a default prior on branch lengths in current Bayesian phylogenetic programs. The default prior in MrBayes assigns independent and identical distributions to branch lengths, imposing strong (and unreasonable) assumptions about the tree length. The problem is exacerbated by the strong correlation between the branch lengths and parameters in models of variable rates among sites or among site partitions. To resolve the problem, we suggest two multivariate priors for the branch lengths (called compound Dirichlet priors) that are fairly diffuse and demonstrate their utility in the special case of branch length estimation on a star phylogeny. Our analysis highlights the need for careful thought in the specification of high-dimensional priors in Bayesian analyses.

Don't throw out the Bayes with the bathwater

AbstractWe highlight one way in which Jones &amp; Love (J&amp;L) misconstrue the Bayesian program: Bayesian models do not represent a rejection of mechanism. This mischaracterization obscures the valid criticisms in their article. We conclude that computational-level Bayesian modeling should not be rejected or discouraged a priori, but should be held to the same empirical standards as other models.

Pinning down the theoretical commitments of Bayesian cognitive models

AbstractMathematical developments in probabilistic inference have led to optimism over the prospects for Bayesian models of cognition. Our target article calls for better differentiation of these technical developments from theoretical contributions. It distinguishes between Bayesian Fundamentalism, which is theoretically limited because of its neglect of psychological mechanism, and Bayesian Enlightenment, which integrates rational and mechanistic considerations and is thus better positioned to advance psychological theory. The commentaries almost uniformly agree that mechanistic grounding is critical to the success of the Bayesian program. Some commentaries raise additional challenges, which we address here. Other commentaries claim that all Bayesian models are mechanistically grounded, while at the same time holding that they should be evaluated only on a computational level. We argue this contradictory stance makes it difficult to evaluate a model's scientific contribution, and that the psychological commitments of Bayesian models need to be made more explicit.

On Identifying the Optimal Number of Population Clusters via the Deviance Information Criterion

Inferring population structure using Bayesian clustering programs often requires a priori specification of the number of subpopulations, , from which the sample has been drawn. Here, we explore the utility of a common Bayesian model selection criterion, the Deviance Information Criterion (DIC), for estimating . We evaluate the accuracy of DIC, as well as other popular approaches, on datasets generated by coalescent simulations under various demographic scenarios. We find that DIC outperforms competing methods in many genetic contexts, validating its application in assessing population structure.

Assessing the prevalence of hybridization between sympatric Canis species surrounding the red wolf (Canis rufus) recovery area in North Carolina

Predicting spatial patterns of hybridization is important for evolutionary and conservation biology yet are hampered by poor understanding of how hybridizing species can interact. This is especially pertinent in contact zones where hybridizing populations are sympatric. In this study, we examined the extent of red wolf (Canis rufus) colonization and introgression where the species contacts a coyote (C. latrans) population in North Carolina, USA. We surveyed 22,000km(2) in the winter of 2008 for scat and identified individual canids through genetic analysis. Of 614 collected scats, 250 were assigned to canids by mitochondrial DNA (mtDNA) sequencing. Canid samples were genotyped at 6-17 microsatellite loci (nDNA) and assigned to species using three admixture criteria implemented in two Bayesian clustering programs. We genotyped 82 individuals but none were identified as red wolves. Two individuals had red wolf mtDNA but no significant red wolf nDNA ancestry. One individual possessed significant red wolf nDNA ancestry (approximately 30%) using all criteria, although seven other individuals showed evidence of red wolf ancestry (11-21%) using the relaxed criterion. Overall, seven individuals were classified as hybrids using the conservative criteria and 37 using the relaxed criterion. We found evidence of dog (C. familiaris) and gray wolf (C. lupus) introgression into the coyote population. We compared the performance of different methods and criteria by analyzing known red wolves and hybrids. These results suggest that red wolf colonization and introgression in North Carolina is minimal and provide insights into the utility of Bayesian clustering methods to detect hybridization.

Pharmacokinetics and dose requirements of disopyramide in neonates and children

Abstract The pharmacokinetics and dose requirements of disopyramide were evaluated in 10 infants and children during the course of routine therapeutic drug monitoring. Drug clearance was estimated using a Bayesian parameter estimation program and found to range from 0.09-0.31L/h/kg. Concentrations associated with a satisfactory response were achieved with doses of 15-20mg/kg/day in most patients, but some required higher doses of 25-30mg/kg/day.