Bayesian inferences in the Cox model for order-restricted hypotheses

In studying the relationship between an ordered categorical predictor and an event time, it is standard practice to include dichotomous indicators of the different levels of the predictor in a Cox model. One can then use a multiple degree-of-freedom score or partial likelihood ratio test for hypothesis testing. Often, interest focuses on comparing the null hypothesis of no difference to an order-restricted alternative, such as a monotone increase across levels of a predictor. This article proposes a Bayesian approach for addressing hypotheses of this type. We reparameterize the Cox model in terms of a cumulative product of parameters having conjugate prior densities, consisting of mixtures of point masses at one, and truncated gamma densities. Due to the structure of the model, posterior computation can proceed via a simple and efficient Gibbs sampling algorithm. Posterior probabilities for the global null hypothesis and subhypotheses, comparing the hazards for specific groups, can be calculated directly from the output of a single Gibbs chain. The approach allows for level sets across which a predictor has no effect. Generalizations to multiple predictors are described, and the method is applied to a study of emergency medical treatment for stroke.     

Bayesian flux balance analysis applied to a skeletal muscle metabolic model

In this article, the steady state condition for the multi-compartment models for cellular metabolism is considered. The problem is to estimate the reaction and transport fluxes, as well as the concentrations in venous blood when the stoichiometry and bound constraints for the fluxes and the concentrations are given. The problem has been addressed previously by a number of authors, and optimization-based approaches as well as extreme pathway analysis have been proposed. These approaches are briefly discussed here. The main emphasis of this work is a Bayesian statistical approach to the flux balance analysis (FBA). We show how the bound constraints and optimality conditions such as maximizing the oxidative phosphorylation flux can be incorporated into the model in the Bayesian framework by proper construction of the prior densities. We propose an effective Markov chain Monte Carlo (MCMC) scheme to explore the posterior densities, and compare the results with those obtained via the previously studied linear programming (LP) approach. The proposed methodology, which is applied here to a two-compartment model for skeletal muscle metabolism, can be extended to more complex models.     

Relationship between sampling duration and concentration of culturable airborne mould and bacteria on selected culture media

AIMS: This study was designed to evaluate potential effects of sampling duration on observed concentrations of airborne culturable mould and bacteria on selected media. METHODS AND RESULTS: Airborne culturable mould and bacteria from lightly to moderately contaminated environments were collected on selected culture media using two co-located, concurrently operated, Andersen N-6 samplers for five sampling durations in the range of 1-10 min. Differences in mean concentrations, as well as linear relationships between sampling duration and both concentration and variability, were evaluated using nonparametric procedures. For the five sampling durations, there were no significant differences in mean concentrations of mould; for bacteria, there were significant differences, with a trend of decreasing concentrations as sampling duration increased. Data variability decreased with increasing sampling duration for both mould and bacteria. CONCLUSIONS: Airborne culturable mould concentrations were similar for sampling durations in the range of 1-10 min. Airborne bacteria concentrations tended to trend downwards with sampling durations exceeding 3 min. SIGNIFICANCE AND IMPACT OF THE STUDY: This study has shown that sampling durations of 1-10 min are appropriate for collection of airborne culturable mould on malt extract agar (MEA) and dichloran glycerol agar (DG-18); based on the apparent trend of decreasing bacterial sample concentrations associated with increasing sampling duration, sampling durations of 相似文献   

基于分段线性模板先验的正电子成像重建

正电子发射成像的噪声性能较其它的一些成像方法要差得多,为了提高重建结果的分辨率和噪声特性,一般采用Bayesian重建。Bayesian方法需要选择恰当的先验,这种先验既要能够抑止重建结果的噪声,又要能够保留图像密度的突变信息。分段线性模板图像模型利用从其它模态的形态学成像得到的组织结构信息,构造适合要求的先验分布函数,由于采用的先验函数是非凸的并包含超验参数,一般的优化方法难以处理,采用动态后验模拟的方法可以很好地解决这些问题。     

ESTIMATING POPULATION SIZE IN ASYNCHRONOUS AGGREGATIONS: A BAYESIAN APPROACH AND TEST WITH ELEPHANT SEAL CENSUSES

Many organisms reproduce in temporary aggregations where estimates of colony size can be made by direct counts. When individuals are not synchronous, however, early breeders depart before the last arrive, so counts underestimate the total breeding population. We present a model describing a colony's census as a function of arrival, breeding tenure, and the correlation between them, and we use it to illustrate how variance in arrival and tenure affect the census. Counts of breeding female northern elephant seals ( Mirounga angustirostris ) from 1975 to 2007 were used to test the model. Four of the model's parameters—population size, mean and variance of arrival date, and the correlation between arrival date and breeding tenure—could be estimated from census data using a Bayesian approach; prior estimates of two other parameters—mean tenure and its variance—had to be used to avoid overparameterization. The model's predictions fit observed censuses well and produced reliable estimates of population size and arrival behavior, showing that the maximum census was 8%–16% below the total number of breeding females. This method could be used for estimating abundance in any asynchronous aggregation, given independent information on one of the defining distributions: arrival, tenure, or departure.     

Health-related monitoring and assessment of airborne particulate matter

Since 1992, the International Atomic Energy Agency (IAEA) has been promoting studies of air pollution using a standard design of air sampler that provides separation on filters into two size fractions with cutoffs of 2.5 and 10 Μm (approximately). These are the size ranges presently considered to have the most important health consequences. Such filter samples are highly amenable to analysis using nuclear and related techniques. After reviewing some of the health effects of airborne particulate matter and current air quality standards and guidelines, this article provides an overview of current and recent IAEA programs in this area, which involve collaborative activities with participants in more than 40 countries.     

Comparison of the Rotorod to other air samplers for the determination of Ambrosia artemisiifolia pollen concentrations conducted in the Environmental Exposure Unit

The Environmental Exposure Unit (EEU) is a 924 m³ facility (Kingston General Hospital, Ontario) in which uniform concentrations of various pollens in HEPA-filtered air at known rates of laminar airflow can be maintained. This facility provided a unique opportunity to compare several air samplers without the environmental variation inherent in outdoor comparisons. The purpose of this study was to conduct a quantitative comparison of pollen measurements using the Rotorod, Burkard™ Personal Volumetric Air Sampler, Air-O-Cell™ and a 37 mm open-faced filter cassette with a microporous filter in the EEU. Pollen samples were taken during clinical trials being conducted in the Unit. Raw pollen counts/m³ obtained using the different methods were corrected using published particle collection efficiencies for the particle size (∼ ∼20 μm) and airflow. Data were analyzed by ANOVA/Tukey HSD. No statistically significant differences were found between pollen concentrations determined by Rotorod, Air-O-Cell and filter cassette. Pollen levels determined by the Burkard were up to 2 times higher than the other sampling methods. Relative standard deviations were similar for the Rotorod, Burkard, and filter cassette and higher for the Air-O-Cell. This study demonstrated that, under our conditions, the Rotorod sampler provides consistent and reliable measurements of ragweed pollen concentrations.     

Performance of the Coriolis air sampler,a high-volume aerosol-collection system for quantification of airborne spores and pollen grains

The Coriolis δ air sampler manufactured by Bertin Technologies (France) is a continuous air sampler, dedicated to outdoor monitoring of airborne spores and pollen grains. This high-volume sampler is based on patented Coriolis technology delivering a liquid sample. The air is drawn into a conical vial in a whirling type motion using suction; particles are pulled against the wall by centrifugal force. Airborne particles are separated from the air and collected in a liquid medium. This innovative solution allows rapid analysis by several techniques including PCR assay and serological assay in order to measure the antigenicity/allergenicity of pollen grains and fungal spores. Also, traditional counting of pollen grains or taxa identification by optical microscopy can be done. A study has been carried out by the Health Protection Agency (HPA), Porton Down, UK, to measure the physical efficiency of the Coriolis air sampler. The physical efficiency of the sampler for collection of micro-organism-laden particles of various sizes has been compared with that of membrane filter samplers using the techniques described by ISO 14698-1. The Coriolis was operated simultaneously with membrane filter samplers in a controlled room where they were challenged with uniform-sized particles of different diameters containing bacterial spores. For the larger particle sizes, it was found that the physical efficiency of the Coriolis was 92% for 10-μm particles. The biological performance of the Coriolis in the collection of airborne fungal spores and pollen grains was evaluated in comparison with a Hirst spore trap (one-week tape-on-drum type sampler) which is one of the most frequently used traps in the measurement of outdoor pollen grain concentrations. The advantages and limitations of both technologies are discussed. The Coriolis was operated simultaneously with a Hirst spore trap in the sampling station of Réseau National de Surveillance Aérobiologique, France (RNSA); the pollen grain and fungal spore counts were analysed by optical microscopy. The pollen grain count m⁻³ collected was compared for both devices. The dispersion values were obtained and statistical analysis was carried out. This study shows that the Coriolis air sampler provided equivalent recovery of pollen grain and fungal spores compared with the volumetric trap standard method (not significantly different, W test, α = 0.05). Nowadays, the French-led project, acronym MONALISA, with financial support from the European Commission––Life-Environment (LIFE05 ENV/F/000068), is testing this innovative air sampler in order to measure the antigenicity/allergenicity of the main aeroallergen particles, i.e. Betula (birch), Poaceae (grasses), Parietaria (pellitory), Olea spp (olive tree), and Artemisia (mugwort) pollen grains, and Alternaria (fungal spores) to validate a new approach of monitoring instead of quantifying pollen grains by their morphology. The robustness and efficiency of the MONALISA system is being demonstrated at a national level throughout Europe in eight different countries with different bio-climatic and topography characteristics: France, UK, Finland, Poland, Spain, Portugal, Switzerland, and Italy.     

Linear and nonlinear mixed-effects models for censored HIV viral loads using normal/independent distributions

HIV RNA viral load measures are often subjected to some upper and lower detection limits depending on the quantification assays. Hence, the responses are either left or right censored. Linear (and nonlinear) mixed-effects models (with modifications to accommodate censoring) are routinely used to analyze this type of data and are based on normality assumptions for the random terms. However, those analyses might not provide robust inference when the normality assumptions are questionable. In this article, we develop a Bayesian framework for censored linear (and nonlinear) models replacing the Gaussian assumptions for the random terms with normal/independent (NI) distributions. The NI is an attractive class of symmetric heavy-tailed densities that includes the normal, Student's-t, slash, and the contaminated normal distributions as special cases. The marginal likelihood is tractable (using approximations for nonlinear models) and can be used to develop Bayesian case-deletion influence diagnostics based on the Kullback-Leibler divergence. The newly developed procedures are illustrated with two HIV AIDS studies on viral loads that were initially analyzed using normal (censored) mixed-effects models, as well as simulations.