Log-modified Weibull regression models with censored data: Sensitivity and residual analysis

This paper proposes a regression model considering the modified Weibull distribution. This distribution can be used to model bathtub-shaped failure rate functions. Assuming censored data, we consider maximum likelihood and Jackknife estimators for the parameters of the model. We derive the appropriate matrices for assessing local influence on the parameter estimates under different perturbation schemes and we also present some ways to perform global influence. Besides, for different parameter settings, sample sizes and censoring percentages, various simulations are performed and the empirical distribution of the modified deviance residual is displayed and compared with the standard normal distribution. These studies suggest that the residual analysis usually performed in normal linear regression models can be straightforwardly extended for a martingale-type residual in log-modified Weibull regression models with censored data. Finally, we analyze a real data set under log-modified Weibull regression models. A diagnostic analysis and a model checking based on the modified deviance residual are performed to select appropriate models. (c) 2008 Elsevier B.V. All rights reserved.

General results for the beta-modified Weibull distribution

We study in detail the so-called beta-modified Weibull distribution, motivated by the wide use of the Weibull distribution in practice, and also for the fact that the generalization provides a continuous crossover towards cases with different shapes. The new distribution is important since it contains as special sub-models some widely-known distributions, such as the generalized modified Weibull, beta Weibull, exponentiated Weibull, beta exponential, modified Weibull and Weibull distributions, among several others. It also provides more flexibility to analyse complex real data. Various mathematical properties of this distribution are derived, including its moments and moment generating function. We examine the asymptotic distributions of the extreme values. Explicit expressions are also derived for the chf, mean deviations, Bonferroni and Lorenz curves, reliability and entropies. The estimation of parameters is approached by two methods: moments and maximum likelihood. We compare by simulation the performances of the estimates from these methods. We obtain the expected information matrix. Two applications are presented to illustrate the proposed distribution.

Double generalized linear model for tissue culture proportion data: a Bayesian perspective

Joint generalized linear models and double generalized linear models (DGLMs) were designed to model outcomes for which the variability can be explained using factors and/or covariates. When such factors operate, the usual normal regression models, which inherently exhibit constant variance, will under-represent variation in the data and hence may lead to erroneous inferences. For count and proportion data, such noise factors can generate a so-called overdispersion effect, and the use of binomial and Poisson models underestimates the variability and, consequently, incorrectly indicate significant effects. In this manuscript, we propose a DGLM from a Bayesian perspective, focusing on the case of proportion data, where the overdispersion can be modeled using a random effect that depends on some noise factors. The posterior joint density function was sampled using Monte Carlo Markov Chain algorithms, allowing inferences over the model parameters. An application to a data set on apple tissue culture is presented, for which it is shown that the Bayesian approach is quite feasible, even when limited prior information is available, thereby generating valuable insight for the researcher about its experimental results.

Modelling the potential distribution of the invasive tomato red spider mite, Tetranychus evansi (Acari: Tetranychidae)

Predicting the potential geographical distribution of a species is particularly important for pests with strong invasive abilities. Tetranychus evansi Baker & Pritchard, possibly native to South America, is a spider mite pest of solanaceous crops. This mite is considered an invasive species in Africa and Europe. A CLIMEX model was developed to predict its global distribution. The model results fitted the known records of T. evansi except for some records in dry locations. Dryness as well as excess moisture stresses play important roles in limiting the spread of the mite in the tropics. In North America and Eurasia its potential distribution appears to be essentially limited by cold stress. Detailed potential distribution maps are provided for T. evansi in the Mediterranean Basin and in Japan. These two regions correspond to climatic borders for the species. Mite establishment in these areas can be explained by their relatively mild winters. The Mediterranean region is also the main area where tomato is grown in open fields in Europe and where the pest represents a threat. According to the model, the whole Mediterranean region has the potential to be extensively colonized by the mite. Wide expansion of the mite to new areas in Africa is also predicted. Agricultural issues highlighted by the modelled distribution of the pest are discussed.

Macroscopic root water uptake distribution using a matric flux potential approach

Hydrological models featuring root water uptake usually do not include compensation mechanisms such that reductions in uptake from dry layers are compensated by an increase in uptake from wetter layers. We developed a physically based root water uptake model with an implicit compensation mechanism. Based on an expression for the matric flux potential (M) as a function of the distance to the root, and assuming a depth-independent value of M at the root surface, uptake per layer is shown to be a function of layer bulk M, root surface M, and a weighting factor that depends on root length density and root radius. Actual transpiration can be calculated from the sum of layer uptake rates. The proposed reduction function (PRF) was built into the SWAP model, and predictions were compared to those made with the Feddes reduction function (FRF). Simulation results were tested against data from Canada (continuous spring wheat [(Triticum aestivum L.]) and Germany (spring wheat, winter barley [Hordeum vulgare L.], sugarbeet [Beta vulgaris L.], winter wheat rotation). For the Canadian data, the root mean square error of prediction (RMSEP) for water content in the upper soil layers was very similar for FRF and PRF; for the deeper layers, RMSEP was smaller for PRF. For the German data, RMSEP was lower for PRF in the upper layers and was similar for both models in the deeper layers. In conclusion, but dependent on the properties of the data sets available for testing,the incorporation of the new reduction function into SWAP was successful, providing new capabilities for simulating compensated root water uptake without increasing the number of input parameters or degrading model performance.

Bayesian ratemaking procedure of crop insurance contracts with skewed distribution

Over the years, crop insurance programs became the focus of agricultural policy in the USA, Spain, Mexico, and more recently in Brazil. Given the increasing interest in insurance, accurate calculation of the premium rate is of great importance. We address the crop-yield distribution issue and its implications in pricing an insurance contract considering the dynamic structure of the data and incorporating the spatial correlation in the Hierarchical Bayesian framework. Results show that empirical (insurers) rates are higher in low risk areas and lower in high risk areas. Such methodological improvement is primarily important in situations of limited data.

A Log-Linear Regression Model for the Beta-Weibull Distribution

We introduce the log-beta Weibull regression model based on the beta Weibull distribution (Famoye et al., 2005; Lee et al., 2007). We derive expansions for the moment generating function which do not depend on complicated functions. The new regression model represents a parametric family of models that includes as sub-models several widely known regression models that can be applied to censored survival data. We employ a frequentist analysis, a jackknife estimator, and a parametric bootstrap for the parameters of the proposed model. We derive the appropriate matrices for assessing local influences on the parameter estimates under different perturbation schemes and present some ways to assess global influences. Further, for different parameter settings, sample sizes, and censoring percentages, several simulations are performed. In addition, the empirical distribution of some modified residuals are displayed and compared with the standard normal distribution. These studies suggest that the residual analysis usually performed in normal linear regression models can be extended to a modified deviance residual in the proposed regression model applied to censored data. We define martingale and deviance residuals to evaluate the model assumptions. The extended regression model is very useful for the analysis of real data and could give more realistic fits than other special regression models.

Occurrence and distribution of Vibrio parahaemolyticus in retail oysters in Sao Paulo State, Brazil

Vibrio parahaemolyticus is a potentially pathogenic bacterium that occurs naturally in estuarine environments worldwide, and is often associated with gastroenteritis in humans following consumption of raw bivalve mollusks, especially raw oysters. The occurrence of total and pathogenic V. parahaemolyticus in 74 samples of raw oysters collected in restaurants, supermarkets, groceries and beach huts in Sao Paulo State, was monitored between February 2006 and January 2007. Enumeration of V. parahaemolyticus was performed according to the most probable number (MPN) procedure. Five to ten typical colonies were selected from thiosulfate-citrate-bile salts-sucrose (TCBS) agar plates for confirmation by the presence of the species-specific gene tlh and the virulence genes tdh and trh by multiplex PCR. V. parahaemolyticus was detected in 100% of samples. The densities of total V. parahaemolyticus varied from 1.78 to 6.04 logio (MPN/g), with higher densities being detected in fall and summer, and lower densities in winter (P < 0.05). There was no statistical difference among densities of V parahaemolyticus regarding the site of collection. None of the 1943 V parahaemolyticus isolates contained tdh and/or trh. These data provide information for the assessment of exposure to V. parahaemolyticus in oysters consumed in Sao Paulo, State, Brazil. (C) 2010 Elsevier Ltd. All rights reserved.

Paste Residence Time in a Spouted Bed Dryer. IV: Effect of the Inert Particle Size Distribution

The residence time distribution and mean residence time of a 10% sodium bicarbonate solution that is dried in a conventional spouted bed with inert bodies were measured with the stimulus-response method. Methylene blue was used as a chemical tracer, and the effects of the paste feed mode, size distribution of the inert bodies, and mean particle size on the residence times and dried powder properties were investigated. The results showed that the residence time distributions could be best reproduced with the perfect mixing cell model or N = 1 for the continuous stirred tank reactor in a series model. The mean residence times ranged from 6.04 to 12.90 min and were significantly affected by the factors studied. Analysis of variance on the experimental data showed that mean residence times were affected by the mean diameter of the inert bodies at a significance level of 1% and by the size distribution at a level of 5%. Moreover, altering the paste feed from dripping to pneumatic atomization affected mean residence time at a 5% significance level. The dried powder characteristics proved to be adequate for further industrial manipulation, as demonstrated by the low moisture content, narrow range of particle size, and good flow properties. The results of this research are significant in the study of the drying of heat-sensitive materials because it shows that by simultaneously changing the size distribution and average size of the inert bodies, the mean residence times of a paste can be reduced by half, thus decreasing losses due to degradation.

Sequence analysis, chromosomal distribution and long-range organization show that rapid turnover of new and old pBuM satellite DNA repeats leads to different patterns of variation in seven species of the Drosophila buzzatii cluster

We aimed to study patterns of variation and factors influencing the evolutionary dynamics of a satellite DNA, pBuM, in all seven Drosophila species from the buzzatii cluster (repleta group). We analyzed 117 alpha pBuM-1 (monomer length 190 bp) and 119 composite alpha/beta (370 bp) pBuM-2 repeats and determined the chromosome location and long-range organization on DNA fibers of major sequence variants. Such combined methodologies in the study of satDNAs have been used in very few organisms. In most species, concerted evolution is linked to high copy number of pBuM repeats. Species presenting low-abundance and scattered distributed pBuM repeats did not undergo concerted evolution and maintained part of the ancestral inter-repeat variability. The alpha and alpha/beta repeats colocalized in heterochromatic regions and were distributed on multiple chromosomes, with notable differences between species. High-resolution FISH revealed array sizes of a few kilobases to over 0.7 Mb and mutual arrangements of alpha and alpha/beta repeats along the same DNA fibers, but with considerable changes in the amount of each variant across species. From sequence, chromosomal and phylogenetic data, we could infer that homogenization and amplification events involved both new and ancestral pBuM variants. Altogether, the data on the structure and organization of the pBuM satDNA give insights into genome evolution including mechanisms that contribute to concerted evolution and diversification.

Genotypic distribution of hepatitis C among hepatitis C and HIV co-infected patients in Brazil

Information on hepatitis C virus (HCV) genotypic distribution among HIV-HCV co-infected patients is lacking in Brazil as well as other Latin American countries. The objective of this study was to evaluate the level of exposure to different risk factors associated with HCV transmission among a group of co-infected patients and to characterize the genotypic distribution of HCV in this cluster. A series of 100 HIV-HCV co-infected patients was analysed. The data to be analysed were collected from specific laboratory tests. Information was collected through a questionnaire. HCV genotyping was carried out by sequencing the 5 ` non-coding region of HCV. Chi-square and Fischer association tests or Kruskal-Wallis test were used to study the association between HCV transmission-related variables and the established genotypes. In conclusion, exposure to multiple risk factors associated with HCV transmission was common among HIV co-infected patients and an association between HCV genotype 3 and intravenous drug user was observed.

Posttranslational mechanisms associated with reduced NHE3 activity in adult vs. young prehypertensive SHR

Crajoinas RO, Lessa LMA, Carraro-Lacroix LR, Davel APC, Pacheco BPM, Rossoni LV, Malnic G, Girardi ACC. Posttranslational mechanisms associated with reduced NHE3 activity in adult vs. young prehypertensive SHR. Am J Physiol Renal Physiol 299:F872-F881, 2010. First published July 14, 2010; doi:10.1152/ajprenal.00654.2009.-Abnormalities in renal proximal tubular (PT) sodium transport play an important role in the pathophysiology of essential hypertension. The Na(+)/H(+) exchanger isoform 3 (NHE3) represents the major route for sodium entry across the apical membrane of renal PT cells. We therefore aimed to assess in vivo NHE3 transport activity and to define the molecular mechanisms underlying NHE3 regulation before and after development of hypertension in the spontaneously hypertensive rat (SHR). NHE3 function was measured as the rate of bicarbonate reabsorption by means of in vivo stationary microperfusion in PT from young prehypertensive SHR (Y-SHR; 5-wk-old), adult SHR (A-SHR; 14-wk-old), and age-matched Wistar Kyoto (WKY) rats. We found that NHE3-mediated PT bicarbonate reabsorption was reduced with age in the SHR (1.08 +/- 0.10 vs. 0.41 +/- 0.04 nmol/cm(2)xs), while it was increased in the transition from youth to adulthood in the WKY rat (0.59 +/- 0.05 vs. 1.26 +/- 0.11 nmol/cm(2)xs). Higher NHE3 activity in the Y-SHR compared with A-SHR was associated with a predominant microvilli confinement and a lower ratio of phosphorylated NHE3 at serine-552 to total NHE3 (P-NHE3/total). After development of hypertension, P-NHE3/total increased and NHE3 was retracted out of the microvillar microdomain along with the regulator dipeptidyl peptidase IV (DPPIV). Collectively, our data suggest that the PT is playing a role in adapting to the hypertension in the SHR. The molecular mechanisms of this adaptation possibly include an increase of P-NHE3/total and a redistribution of the NHE3-DPPIV complex from the body to the base of the PT microvilli, both predicted to decrease sodium reabsorption.

Prednisone monotherapy induced remission in a group of patients with membranous lupus nephritis

The treatment of membranous lupus nephritis (MLN) is still controversial in the literature. We conducted a retrospective analysis of patients in two medical centers of Sao Paulo-Brazil in order to evaluate the clinical response in patients submitted to either a regimen with prednisone alone or to a double immunosuppressive regimen (prednisone plus cyclophosphamide or prednisone plus azathioprine). Methods: MLN female patients were enrolled in this retrospective study conducted from February 1999 to June 2007. Data were collected from the patients` medical charts. Race distribution was similar in both groups: Caucasian (72.3%) and Afro-Latin-American (27.7%). The prednisone regimen consisted of 1 mg/kg/day for 8 weeks and tapering until 0.1 mg/kg/day (n = 29). The double immunosuppressive treatment consisted of the same doses of prednisone plus monthly intravenous cyclophosphamide or azathioprine for 6 months (n = 24). Criteria for remission (complete and partial) and renal function loss as well as flare criteria followed those used in the literature. Results: There was no difference between the prednisone group and the double immunosuppressive group regarding age (33.2 +/- 9.4 vs. 29.1 +/- 9.1 y), estimated GFR (76.5 +/- 26.6 vs. 74.1 +/- 39.6 ml/min/1.73 m(2)), serum albumin (2.8 +/- 0.7 vs. 2.6 +/- 0.3 g/dl), positive ANA (87.5 vs. 90.0%), positive anti-dsDNA (47.6 vs. 44.0%), renal SLEDAI indices (6.6 +/- 2.6 vs. 7.0 +/- 3.1), follow-up time (71 +/- 46 vs. 62 +/- 45 months), as well as proteinuria (3.1 +/- 1.9 vs. 4.8 +/- 2.4 g/day) and number of non-nephrotic patients (6 in the prednisone group vs. 3 in the double immunosuppressive group). The prednisone group presented higher C3 values (85.2 +/- 31.5 vs. 62.3 +/- 41.6 U/ml, p = 0.04). Clinical and laboratory characteristics at 6 months and at last follow-up did not reveal any differences between treatment regimens. Renal survival after an 8-year follow-up did not differ in both groups (prednisone group 86.2% vs. double immunosuppressive group 75%), and patients in both groups showed a high rate of renal flares (prednisone group 51.7% vs. double immunosuppressive group 62.5%). Univariate analysis showed that only patient age predicted flares (r = -0.048, p = 0.04). Borderline significance was obtained for proteinuria analysis (p = 0.07). Adverse effects did not differ between the groups. Conclusions: A regimen of corticosteroids in MLN induced a high remission rate after 6 months. Both treatment regimens showed a high flare rate and age was the only predictive parameter (r = -0.048, p = 0.04). Renal survival after 8 years did not differ between the groups.

Initial hepatosplanchnic blood flow distribution and oxygen metabolism in experimental model of hypotensive brain death

Background: Organs from the so-called marginal donors have been used with a significant higher risk of primary non function than organs retrieved from the optimal donors. We investigated the early metabolic changes and blood flow redistribution in splanchnic territory in an experimental model that mimics marginal brain-dead (BD) donor. Material/Methods: Ten dogs (21.3 +/- 0.9 kg), were subjected to a brain death protocol induced by subdural balloon inflation and observed for 30 min thereafter without ally additional interventions. Mean arterial and intracranial pressures, heart rate, cardiac output (CO), portal vein and hepatic artery blood flows (PVBF and HABF, ultrasonic flowprobe), and O(2)-derived variables were evaluated. Results: An increase in arterial pressure, CO, PVBF and HABF was observed after BD induction. At the end, an intense hypotension with normalization in CO (3.0 +/- 0.2 VS. 2.8 +/- 2.8 L/min) and PVBF (687 +/- 114 vs. 623 +/- 130 ml/min) was observed, whereas HABF (277 33 vs. 134 28 ml/min, p<0.005) remained lower than baseline values. Conclusions: Despite severe hypotension induced by sudden increase of intracranial pressure, the systemic and splanchnic blood flows were partially preserved without signs of severe hypoperfusion (i.e. hyperlactatemia). Additionally, the HABF was mostly negatively affected in this model of marginal BD donor. Our data suggest that not only the cardiac output, but the intrinsic hepatic microcirculatory mechanism plays a role in the hepatic blood flow control after BD.

Capnography reflects ventilation/perfusion distribution in a model of acute lung injury

Background Changes in the shape of the capnogram may reflect changes in lung physiology. We studied the effect of different ventilation/perfusion ratios (V/Q) induced by positive end-expiratory pressures (PEEP) and lung recruitment on phase III slope (S(III)) of volumetric capnograms. Methods Seven lung-lavaged pigs received volume control ventilation at tidal volumes of 6 ml/kg. After a lung recruitment maneuver, open-lung PEEP (OL-PEEP) was defined at 2 cmH(2)O above the PEEP at the onset of lung collapse as identified by the maximum respiratory compliance during a decremental PEEP trial. Thereafter, six distinct PEEP levels either at OL-PEEP, 4 cmH(2)O above or below this level were applied in a random order, either with or without a prior lung recruitment maneuver. Ventilation-perfusion distribution (using multiple inert gas elimination technique), hemodynamics, blood gases and volumetric capnography data were recorded at the end of each condition (minute 40). Results S(III) showed the lowest value whenever lung recruitment and OL-PEEP were jointly applied and was associated with the lowest dispersion of ventilation and perfusion (Disp(R-E)), the lowest ratio of alveolar dead space to alveolar tidal volume (VD(alv)/VT(alv)) and the lowest difference between arterial and end-tidal pCO(2) (Pa-ETCO(2)). Spearman`s rank correlations between S(III) and Disp(R-E) showed a =0.85 with 95% CI for (Fisher`s Z-transformation) of 0.74-0.91, P < 0.0001. Conclusion In this experimental model of lung injury, changes in the phase III slope of the capnograms were directly correlated with the degree of ventilation/perfusion dispersion.