STOCHASTIC MODEL FOR SIMULATING MAIZE YIELD

Maize is one of the most important crops in the world. The products generated from this crop are largely used in the starch industry, the animal and human nutrition sector, and biomass energy production and refineries. For these reasons, there is much interest in figuring the potential grain yield of maize genotypes in relation to the environment in which they will be grown, as the productivity directly affects agribusiness or farm profitability. Questions like these can be investigated with ecophysiological crop models, which can be organized according to different philosophies and structures. The main objective of this work is to conceptualize a stochastic model for predicting maize grain yield and productivity under different conditions of water supply while considering the uncertainties of daily climate data. Therefore, one focus is to explain the model construction in detail, and the other is to present some results in light of the philosophy adopted. A deterministic model was built as the basis for the stochastic model. The former performed well in terms of the curve shape of the above-ground dry matter over time as well as the grain yield under full and moderate water deficit conditions. Through the use of a triangular distribution for the harvest index and a bivariate normal distribution of the averaged daily solar radiation and air temperature, the stochastic model satisfactorily simulated grain productivity, i.e., it was found that 10,604 kg ha(-1) is the most likely grain productivity, very similar to the productivity simulated by the deterministic model and for the real conditions based on a field experiment.

Evaluation of Viral Load Thresholds for Predicting New World Health Organization Stage 3 and 4 Events in HIV-Infected Children Receiving Highly Active Antiretroviral Therapy

Background: This study evaluated a wide range of viral load (VL) thresholds to identify a cut-point that best predicts new clinical events in children on stable highly active antiretroviral therapy (HAART). Methods: Cox proportional hazards modeling was used to assess the adjusted risk for World Health Organization stage 3 or 4 clinical events (WHO events) as a function of time-varying CD4, VL, and hemoglobin values in a cohort study of Latin American children on HAART >= 6 months. Models were fit using different VL cut-points between 400 and 50,000 copies per milliliter, with model fit evaluated on the basis of the minimum Akaike information criterion value, a standard model fit statistic. Results: Models were based on 67 subjects with WHO events out of 550 subjects on study. The VL cut-points of >2600 and >32,000 copies per milliliter corresponded to the lowest Akaike information criterion values and were associated with the highest hazard ratios (2.0, P = 0.015; and 2.1, P = 0.0058, respectively) for WHO events. Conclusions: In HIV-infected Latin American children on stable HAART, 2 distinct VL thresholds (>2600 and >32,000 copies/mL) were identified for predicting children at significantly increased risk for HIV-related clinical illness, after accounting for CD4 level, hemoglobin level, and other significant factors.

Predicting smear negative pulmonary tuberculosis with classification trees and logistic regression: a cross-sectional study

Abstract Background Smear negative pulmonary tuberculosis (SNPT) accounts for 30% of pulmonary tuberculosis cases reported yearly in Brazil. This study aimed to develop a prediction model for SNPT for outpatients in areas with scarce resources. Methods The study enrolled 551 patients with clinical-radiological suspicion of SNPT, in Rio de Janeiro, Brazil. The original data was divided into two equivalent samples for generation and validation of the prediction models. Symptoms, physical signs and chest X-rays were used for constructing logistic regression and classification and regression tree models. From the logistic regression, we generated a clinical and radiological prediction score. The area under the receiver operator characteristic curve, sensitivity, and specificity were used to evaluate the model's performance in both generation and validation samples. Results It was possible to generate predictive models for SNPT with sensitivity ranging from 64% to 71% and specificity ranging from 58% to 76%. Conclusion The results suggest that those models might be useful as screening tools for estimating the risk of SNPT, optimizing the utilization of more expensive tests, and avoiding costs of unnecessary anti-tuberculosis treatment. Those models might be cost-effective tools in a health care network with hierarchical distribution of scarce resources.

Validation of models for predicting milk urea nitrogen concentrations, estimating dry matter intake by the NRC (2001)

The objective of this study was to validate three different models for predicting milk urea nitrogen using field conditions, attempting to evaluate the nutritional adequacy diets for dairy cows and prediction of nitrogen excreted to the environment. Observations (4,749) from 855 cows were used. Milk yield, body weight (BW), days in milk and parity were recorded on the milk sampling days. Milk was sampled monthly, for analysis of milk urea nitrogen (MUN), fat, protein, lactose and total solids concentration and somatic cells count. Individual dry matter intake was estimated using the NRC (2001). The three models studied were derived from a first one to predict urinary nitrogen (UN). Model 1 was MUN = UN/12.54, model 2 was MUN = UN/17.6 and model 3 was MUN = UN/(0.0259 × BW), adjusted by body weight effect. To evaluate models, they were tested for accuracy, precision and robustness. Despite being more accurate (mean bias = 0.94 mg/dL), model 2 was less precise (residual error = 4.50 mg/dL) than model 3 (mean bias = 1.41 and residual error = 4.11 mg/dL), while model 1 was the least accurate (mean bias = 6.94 mg/dL) and the least precise (residual error = 5.40 mg/dL). They were not robust, because they were influenced by almost all the variables studied. The three models for predicting milk urea nitrogen were different with respect to accuracy, precision and robustness.

Two fluid numerical model for coastal applications

Wave breaking is an important coastal process, influencing hydro-morphodynamic processes such as turbulence generation and wave energy dissipation, run-up on the beach and overtopping of coastal defence structures. During breaking, waves are complex mixtures of air and water (“white water”) whose properties affect velocity and pressure fields in the vicinity of the free surface and, depending on the breaker characteristics, different mechanisms for air entrainment are usually observed. Several laboratory experiments have been performed to investigate the role of air bubbles in the wave breaking process (Chanson & Cummings, 1994, among others) and in wave loading on vertical wall (Oumeraci et al., 2001; Peregrine et al., 2006, among others), showing that the air phase is not negligible since the turbulent energy dissipation involves air-water mixture. The recent advancement of numerical models has given valuable insights in the knowledge of wave transformation and interaction with coastal structures. Among these models, some solve the RANS equations coupled with a free-surface tracking algorithm and describe velocity, pressure, turbulence and vorticity fields (Lara et al. 2006 a-b, Clementi et al., 2007). The single-phase numerical model, in which the constitutive equations are solved only for the liquid phase, neglects effects induced by air movement and trapped air bubbles in water. Numerical approximations at the free surface may induce errors in predicting breaking point and wave height and moreover, entrapped air bubbles and water splash in air are not properly represented. The aim of the present thesis is to develop a new two-phase model called COBRAS2 (stands for Cornell Breaking waves And Structures 2 phases), that is the enhancement of the single-phase code COBRAS0, originally developed at Cornell University (Lin & Liu, 1998). In the first part of the work, both fluids are considered as incompressible, while the second part will treat air compressibility modelling. The mathematical formulation and the numerical resolution of the governing equations of COBRAS2 are derived and some model-experiment comparisons are shown. In particular, validation tests are performed in order to prove model stability and accuracy. The simulation of the rising of a large air bubble in an otherwise quiescent water pool reveals the model capability to reproduce the process physics in a realistic way. Analytical solutions for stationary and internal waves are compared with corresponding numerical results, in order to test processes involving wide range of density difference. Waves induced by dam-break in different scenarios (on dry and wet beds, as well as on a ramp) are studied, focusing on the role of air as the medium in which the water wave propagates and on the numerical representation of bubble dynamics. Simulations of solitary and regular waves, characterized by both spilling and plunging breakers, are analyzed with comparisons with experimental data and other numerical model in order to investigate air influence on wave breaking mechanisms and underline model capability and accuracy. Finally, modelling of air compressibility is included in the new developed model and is validated, revealing an accurate reproduction of processes. Some preliminary tests on wave impact on vertical walls are performed: since air flow modelling allows to have a more realistic reproduction of breaking wave propagation, the dependence of wave breaker shapes and aeration characteristics on impact pressure values is studied and, on the basis of a qualitative comparison with experimental observations, the numerical simulations achieve good results.

Aerosol-cloud interactions studied with a chemistry-climate model

This study aims at a comprehensive understanding of the effects of aerosol-cloud interactions and their effects on cloud properties and climate using the chemistry-climate model EMAC. In this study, CCN activation is regarded as the dominant driver in aerosol-cloud feedback loops in warm clouds. The CCN activation is calculated prognostically using two different cloud droplet nucleation parameterizations, the STN and HYB CDN schemes. Both CDN schemes account for size and chemistry effects on the droplet formation based on the same aerosol properties. The calculation of the solute effect (hygroscopicity) is the main difference between the CDN schemes. The kappa-method is for the first time incorporated into Abdul-Razzak and Ghan activation scheme (ARG) to calculate hygroscopicity and critical supersaturation of aerosols (HYB), and the performance of the modied scheme is compared with the osmotic coefficient model (STN), which is the standard in the ARG scheme. Reference simulations (REF) with the prescribed cloud droplet number concentration have also been carried out in order to understand the effects of aerosol-cloud feedbacks. In addition, since the calculated cloud coverage is an important determinant of cloud radiative effects and is influencing the nucleation process two cloud cover parameterizations (i.e., a relative humidity threshold; RH-CLC and a statistical cloud cover scheme; ST-CLC) have been examined together with the CDN schemes, and their effects on the simulated cloud properties and relevant climate parameters have been investigated. The distinct cloud droplet spectra show strong sensitivity to aerosol composition effects on cloud droplet formation in all particle sizes, especially for the Aitken mode. As Aitken particles are the major component of the total aerosol number concentration and CCN, and are most sensitive to aerosol chemical composition effect (solute effect) on droplet formation, the activation of Aitken particles strongly contribute to total cloud droplet formation and thereby providing different cloud droplet spectra. These different spectra influence cloud structure, cloud properties, and climate, and show regionally varying sensitivity to meteorological and geographical condition as well as the spatiotemporal aerosol properties (i.e., particle size, number, and composition). The changes responding to different CDN schemes are more pronounced at lower altitudes than higher altitudes. Among regions, the subarctic regions show the strongest changes, as the lower surface temperature amplifies the effects of the activated aerosols; in contrast, the Sahara desert, where is an extremely dry area, is less influenced by changes in CCN number concentration. The aerosol-cloud coupling effects have been examined by comparing the prognostic CDN simulations (STN, HYB) with the reference simulation (REF). Most pronounced effects are found in the cloud droplet number concentration, cloud water distribution, and cloud radiative effect. The aerosol-cloud coupling generally increases cloud droplet number concentration; this decreases the efficiency of the formation of weak stratiform precipitation, and increases the cloud water loading. These large-scale changes lead to larger cloud cover and longer cloud lifetime, and contribute to high optical thickness and strong cloud cooling effects. This cools the Earth's surface, increases atmospheric stability, and reduces convective activity. These changes corresponding to aerosol-cloud feedbacks are also differently simulated depending on the cloud cover scheme. The ST-CLC scheme is more sensitive to aerosol-cloud coupling, since this scheme uses a tighter linkage of local dynamics and cloud water distributions in cloud formation process than the RH-CLC scheme. For the calculated total cloud cover, the RH-CLC scheme simulates relatively similar pattern to observations than the ST-CLC scheme does, but the overall properties (e.g., total cloud cover, cloud water content) in the RH simulations are overestimated, particularly over ocean. This is mainly originated from the difference in simulated skewness in each scheme: the RH simulations calculate negatively skewed distributions of cloud cover and relevant cloud water, which is similar to that of the observations, while the ST simulations yield positively skewed distributions resulting in lower mean values than the RH-CLC scheme does. The underestimation of total cloud cover over ocean, particularly over the intertropical convergence zone (ITCZ) relates to systematic defficiency of the prognostic calculation of skewness in the current set-ups of the ST-CLC scheme.rnOverall, the current EMAC model set-ups perform better over continents for all combinations of the cloud droplet nucleation and cloud cover schemes. To consider aerosol-cloud feedbacks, the HYB scheme is a better method for predicting cloud and climate parameters for both cloud cover schemes than the STN scheme. The RH-CLC scheme offers a better simulation of total cloud cover and the relevant parameters with the HYB scheme and single-moment microphysics (REF) than the ST-CLC does, but is not very sensitive to aerosol-cloud interactions.

Predicting adverse events in children with fever and chemotherapy-induced neutropenia: the prospective multicenter SPOG 2003 FN study

PURPOSE To develop a score predicting the risk of adverse events (AEs) in pediatric patients with cancer who experience fever and neutropenia (FN) and to evaluate its performance. PATIENTS AND METHODS Pediatric patients with cancer presenting with FN induced by nonmyeloablative chemotherapy were observed in a prospective multicenter study. A score predicting the risk of future AEs (ie, serious medical complication, microbiologically defined infection, radiologically confirmed pneumonia) was developed from a multivariate mixed logistic regression model. Its cross-validated predictive performance was compared with that of published risk prediction rules. Results An AE was reported in 122 (29%) of 423 FN episodes. In 57 episodes (13%), the first AE was known only after reassessment after 8 to 24 hours of inpatient management. Predicting AE at reassessment was better than prediction at presentation with FN. A differential leukocyte count did not increase the predictive performance. The score predicting future AE in 358 episodes without known AE at reassessment used the following four variables: preceding chemotherapy more intensive than acute lymphoblastic leukemia maintenance (weight = 4), hemoglobin > or = 90 g/L (weight = 5), leukocyte count less than 0.3 G/L (weight = 3), and platelet count less than 50 G/L (weight = 3). A score (sum of weights) > or = 9 predicted future AEs. The cross-validated performance of this score exceeded the performance of published risk prediction rules. At an overall sensitivity of 92%, 35% of the episodes were classified as low risk, with a specificity of 45% and a negative predictive value of 93%. CONCLUSION This score, based on four routinely accessible characteristics, accurately identifies pediatric patients with cancer with FN at risk for AEs after reassessment.

Predicting the transition from acute to persistent low back pain

BACKGROUND: Most people experience low back pain (LBP) at least once in their lifetime. Only a minority of them go on to develop persistent LBP. However, the socioeconomic costs of persistent LBP significantly exceed the costs of the initial acute LBP episode. AIMS: To identify factors that influence the progression of acute LBP to the persistent state at an early stage. METHODS: Prospective inception cohort study of patients attending a health practitioner for their first episode of acute LBP or recurrent LBP after a pain free period of at least 6 months. Patients were assessed at baseline addressing occupational and psychological factors as well as pain, disability, quality of life and physical activity and followed up at 3, 6, 12 weeks and 6 months. Variables were combined to the three indices 'working condition', 'depression and maladaptive cognitions' and 'pain and quality of life'. RESULTS: The index 'depression and maladaptive cognitions' was found to be a significant baseline predictor for persistent LBP up to 6 months (OR 5.1; 95% CI: 1.04-25.1). Overall predictive accuracy of the model was 81%. CONCLUSIONS: In this study of patients with acute LBP in a primary care setting psychological factors at baseline correlated with a progression to persistent LBP up to 6 months. The benefit of including factors such as 'depression and maladaptive cognition' in screening tools is that these factors can be addressed in primary and secondary prevention.

Predicting the transition from acute to persistent low back pain

BACKGROUND: Most people experience low back pain (LBP) at least once in their lifetime. Only a minority of them go on to develop persistent LBP. However, the socioeconomic costs of persistent LBP significantly exceed the costs of the initial acute LBP episode. AIMS: To identify factors that influence the progression of acute LBP to the persistent state at an early stage. METHODS: Prospective inception cohort study of patients attending a health practitioner for their first episode of acute LBP or recurrent LBP after a pain free period of at least 6 months. Patients were assessed at baseline addressing occupational and psychological factors as well as pain, disability, quality of life and physical activity and followed up at 3, 6, 12 weeks and 6 months. Variables were combined to the three indices 'working condition', 'depression and maladaptive cognitions' and 'pain and quality of life'. RESULTS: The index 'depression and maladaptive cognitions' was found to be a significant baseline predictor for persistent LBP up to 6 months (OR 5.1; 95% CI: 1.04-25.1). Overall predictive accuracy of the model was 81%. CONCLUSIONS: In this study of patients with acute LBP in a primary care setting psychological factors at baseline correlated with a progression to persistent LBP up to 6 months. The benefit of including factors such as 'depression and maladaptive cognition' in screening tools is that these factors can be addressed in primary and secondary prevention.

Ability of the PM3 Quantum Mechanical Method to Model Intermolecular Hydrogen Bonding between Neutral Molecules

The PM3 semiempirical quantum-mechanical method was found to systematically describe intermolecular hydrogen bonding in small polar molecules. PM3 shows charge transfer from the donor to acceptor molecules on the order of 0.02-0.06 units of charge when strong hydrogen bonds are formed. The PM3 method is predictive; calculated hydrogen bond energies with an absolute magnitude greater than 2 kcal mol-' suggest that the global minimum is a hydrogen bonded complex; absolute energies less than 2 kcal mol-' imply that other van der Waals complexes are more stable. The geometries of the PM3 hydrogen bonded complexes agree with high-resolution spectroscopic observations, gas electron diffraction data, and high-level ab initio calculations. The main limitations in the PM3 method are the underestimation of hydrogen bond lengths by 0.1-0.2 for some systems and the underestimation of reliable experimental hydrogen bond energies by approximately 1-2 kcal mol-l. The PM3 method predicts that ammonia is a good hydrogen bond acceptor and a poor hydrogen donor when interacting with neutral molecules. Electronegativity differences between F, N, and 0 predict that donor strength follows the order F > 0 > N and acceptor strength follows the order N > 0 > F. In the calculations presented in this article, the PM3 method mirrors these electronegativity differences, predicting the F-H- - -N bond to be the strongest and the N-H- - -F bond the weakest. It appears that the PM3 Hamiltonian is able to model hydrogen bonding because of the reduction of two-center repulsive forces brought about by the parameterization of the Gaussian core-core interactions. The ability of the PM3 method to model intermolecular hydrogen bonding means reasonably accurate quantum-mechanical calculations can be applied to small biologic systems.

A prediction model for assessing residential radon concentration in Switzerland

Indoor radon is regularly measured in Switzerland. However, a nationwide model to predict residential radon levels has not been developed. The aim of this study was to develop a prediction model to assess indoor radon concentrations in Switzerland. The model was based on 44,631 measurements from the nationwide Swiss radon database collected between 1994 and 2004. Of these, 80% randomly selected measurements were used for model development and the remaining 20% for an independent model validation. A multivariable log-linear regression model was fitted and relevant predictors selected according to evidence from the literature, the adjusted R², the Akaike's information criterion (AIC), and the Bayesian information criterion (BIC). The prediction model was evaluated by calculating Spearman rank correlation between measured and predicted values. Additionally, the predicted values were categorised into three categories (50th, 50th-90th and 90th percentile) and compared with measured categories using a weighted Kappa statistic. The most relevant predictors for indoor radon levels were tectonic units and year of construction of the building, followed by soil texture, degree of urbanisation, floor of the building where the measurement was taken and housing type (P-values <0.001 for all). Mean predicted radon values (geometric mean) were 66 Bq/m³ (interquartile range 40-111 Bq/m³) in the lowest exposure category, 126 Bq/m³ (69-215 Bq/m³) in the medium category, and 219 Bq/m³ (108-427 Bq/m³) in the highest category. Spearman correlation between predictions and measurements was 0.45 (95%-CI: 0.44; 0.46) for the development dataset and 0.44 (95%-CI: 0.42; 0.46) for the validation dataset. Kappa coefficients were 0.31 for the development and 0.30 for the validation dataset, respectively. The model explained 20% overall variability (adjusted R²). In conclusion, this residential radon prediction model, based on a large number of measurements, was demonstrated to be robust through validation with an independent dataset. The model is appropriate for predicting radon level exposure of the Swiss population in epidemiological research. Nevertheless, some exposure misclassification and regression to the mean is unavoidable and should be taken into account in future applications of the model.

Modeling And Predicting The Performance Of Infiltrated Solid Oxide Fuel Cell Anodes

Fuel cells are a topic of high interest in the scientific community right now because of their ability to efficiently convert chemical energy into electrical energy. This thesis is focused on solid oxide fuel cells (SOFCs) because of their fuel flexibility, and is specifically concerned with the anode properties of SOFCs. The anodes are composed of a ceramic material (yttrium stabilized zirconia, or YSZ), and conducting material. Recent research has shown that an infiltrated anode may offer better performance at a lower cost. This thesis focuses on the creation of a model of an infiltrated anode that mimics the underlying physics of the production process. Using the model, several key parameters for anode performance are considered. These are the initial volume fraction of YSZ in the slurry before sintering, the final porosity of the composite anode after sintering, and the size of the YSZ and conducting particles in the composite. The performance measures of the anode, namely percolation threshold and effective conductivity, are analyzed as a function of these important input parameters. Simple two and three-dimensional percolation models are used to determine the conditions at which the full infiltrated anode would be investigated. These more simple models showed that the aspect ratio of the anode has no effect on the threshold or effective conductivity, and that cell sizes of 303 are needed to obtain accurate conductivity values. The full model of the infiltrated anode is able to predict the performance of the SOFC anodes and it can be seen that increasing the size of the YSZ decreases the percolation threshold and increases the effective conductivity at low conductor loadings. Similar trends are seen for a decrease in final porosity and a decrease in the initial volume fraction of YSZ.

Advantages and Applications of Transforming Empirical Model Input Space with Dimensional Models

Model-based calibration of steady-state engine operation is commonly performed with highly parameterized empirical models that are accurate but not very robust, particularly when predicting highly nonlinear responses such as diesel smoke emissions. To address this problem, and to boost the accuracy of more robust non-parametric methods to the same level, GT-Power was used to transform the empirical model input space into multiple input spaces that simplified the input-output relationship and improved the accuracy and robustness of smoke predictions made by three commonly used empirical modeling methods: Multivariate Regression, Neural Networks and the k-Nearest Neighbor method. The availability of multiple input spaces allowed the development of two committee techniques: a 'Simple Committee' technique that used averaged predictions from a set of 10 pre-selected input spaces chosen by the training data and the "Minimum Variance Committee" technique where the input spaces for each prediction were chosen on the basis of disagreement between the three modeling methods. This latter technique equalized the performance of the three modeling methods. The successively increasing improvements resulting from the use of a single best transformed input space (Best Combination Technique), Simple Committee Technique and Minimum Variance Committee Technique were verified with hypothesis testing. The transformed input spaces were also shown to improve outlier detection and to improve k-Nearest Neighbor performance when predicting dynamic emissions with steady-state training data. An unexpected finding was that the benefits of input space transformation were unaffected by changes in the hardware or the calibration of the underlying GT-Power model.

Predicting spike timing of neocortical pyramidal neurons by simple threshold models

Neurons generate spikes reliably with millisecond precision if driven by a fluctuating current--is it then possible to predict the spike timing knowing the input? We determined parameters of an adapting threshold model using data recorded in vitro from 24 layer 5 pyramidal neurons from rat somatosensory cortex, stimulated intracellularly by a fluctuating current simulating synaptic bombardment in vivo. The model generates output spikes whenever the membrane voltage (a filtered version of the input current) reaches a dynamic threshold. We find that for input currents with large fluctuation amplitude, up to 75% of the spike times can be predicted with a precision of +/-2 ms. Some of the intrinsic neuronal unreliability can be accounted for by a noisy threshold mechanism. Our results suggest that, under random current injection into the soma, (i) neuronal behavior in the subthreshold regime can be well approximated by a simple linear filter; and (ii) most of the nonlinearities are captured by a simple threshold process.

Clinical and biochemical description of a novel CYP21A2 gene mutation 962_963insA using a new 3D model for the P450c21 protein

OBJECTIVE: A severely virilized 46, XX newborn girl was referred to our center for evaluation and treatment of congenital adrenal hyperplasia (CAH) because of highly elevated 17alpha-hydroxyprogesterone levels at newborn screening; biochemical tests confirmed the diagnosis of salt-wasting CAH. Genetic analysis revealed that the girl was compound heterozygote for a previously reported Q318X mutation in exon 8 and a novel insertion of an adenine between nucleotides 962 and 963 in exon 4 of the CYP21A2 gene. This 962_963insA mutation created a frameshift leading to a stop codon at amino acid 161 of the P450c21 protein. AIM AND METHODS: To better understand structure-function relationships of mutant P450c21 proteins, we performed multiple sequence alignments of P450c21 with three mammalian P450s (P450 2C8, 2C9 and 2B4) with known structures as well as with human P450c17. Comparative molecular modeling of human P450c21 was then performed by MODELLER using the X-ray crystal structure of rabbit P450 2B4 as a template. RESULTS: The new three dimensional model of human P450c21 and the sequence alignment were found to be helpful in predicting the role of various amino acids in P450c21, especially those involved in heme binding and interaction with P450 oxidoreductase, the obligate electron donor. CONCLUSION: Our model will help in analyzing the genotype-phenotype relationship of P450c21 mutations which have not been tested for their functional activity in an in vitro assay.