Sequential Modeling Approach Research Articles

Sequential dynamic artificial neural network modeling of a full-scale coking wastewater treatment plant with fluidized bed reactors.

This study proposed a sequential modeling approach using an artificial neural network (ANN) to develop four independent models which were able to predict biotreatment effluent variables of a full-scale coking wastewater treatment plant (CWWTP). Suitable structure and transfer function of ANN were optimized by genetic algorithm. The sequential approach, which included two parts, an influent estimator and an effluent predictor, was used to develop dynamic models. The former parts of models estimated the variations of influent COD, volatile phenol, cyanide, and NH4 (+)-N. The later parts of models predicted effluent COD, volatile phenol, cyanide, and NH4 (+)-N using the estimated values and other parameters. The performance of these models was evaluated by statistical parameters (such as coefficient of determination (R (2) ), etc.). Obtained results indicated that the estimator developed dynamic models for influent COD (R (2) = 0.871), volatile phenol (R (2) = 0.904), cyanide (R (2) = 0.846), and NH4 (+)-N (R (2) = 0.777), while the predictor developed feasible models for effluent COD (R (2) = 0.852) and cyanide (R (2) = 0.844), with slightly worse models for effluent volatile phenol (R (2) = 0.752) and NH4 (+)-N (R (2) = 0.764). Thus, the proposed modeling processes can be used as a tool for the prediction of CWWTP performance.

FLT3 and CDK4/6 inhibitors: signaling mechanisms and tumor burden in subcutaneous and orthotopic mouse models of acute myeloid leukemia.

FLT3(ITD) subtype acute myeloid leukemia (AML) has a poor prognosis with currently available therapies. A number of small molecule inhibitors of FLT3 and/or CDK4/6 are currently under development. A more complete and quantitative understanding of the mechanisms of action of FLT3 and CDK4/6 inhibitors may better inform the development of current and future compounds that act on one or both of the molecular targets, and thus may lead to improved treatments for AML. In this study, we investigated in both subcutaneous and orthotopic AML mouse models, the mechanisms of action of three FLT3 and/or CDK4/6 inhibitors: AMG925 (Amgen), sorafenib (Bayer and Onyx), and quizartinib (Ambit Biosciences). A composite model was developed to integrate the plasma pharmacokinetics of these three compounds on their respective molecular targets, the coupling between the target pathways, as well as the resulting effects on tumor burden reduction in the subcutaneous xenograft model. A sequential modeling approach was used, wherein model structures and estimated parameters from upstream processes (e.g. PK, cellular signaling) were fixed for modeling subsequent downstream processes (cellular signaling, tumor burden). Pooled data analysis was employed for the plasma PK and cellular signaling modeling, while population modeling was applied to the tumor burden modeling. The resulting model allows the decomposition of the relative contributions of FLT3(ITD) and CDK4/6 inhibition on downstream signaling and tumor burden. In addition, the action of AMG925 on cellular signaling and tumor burden was further studied in an orthotopic tumor mouse model more closely representing the physiologically relevant environment for AML.

Bi-level decisions of vacant taxi drivers traveling towards taxi stands in customer-search: Modeling methodology and policy implications

This study adopts the sequential logit approach to modeling bi-level decisions of vacant taxi drivers in customer-search. The first level decision is about whether the drivers will travel to one of the nearest taxi stands after dropping off their customers and the second level decision is on whether the drivers will join the queue at the nearest taxi stand once they have arrived there. A stated preference survey was conducted to interview 258 urban taxi drivers about their choices of the two level decisions. The statistical test shows that search districts, travel distance from the customer׳s drop-off location to the designated taxi stand, the congestion level on the way of cruising, as well as the preference for traveling towards taxi stands are found to be the significant factors of the first level decision. This study also confirms that the queue lengths of both taxis and passengers at taxi stands, the expected customer-search distance after leaving taxi stands, and the preference of vacant taxi drivers for staying at taxi stands are found to be significantly influence the second level decision. The likelihood ratio tests for market segmentation analysis demonstrate the variations in preferences of taxi drivers operated in different taxi shifts and service areas. Some policy implications on introducing more taxi stands and improving the utilization rates of taxi stands are also discussed. We believe that the proposed sequential logit modeling approach, findings, and discussions are useful for developing micro-simulation models in terms of evaluating the performance of road traffic networks with taxi services and for developing simulation-based optimization models to answer policy questions related to taxi services.

Detecting trends in diameter growth of Norway spruce on long-term forest research plots using linear mixed-effects models

The objective of the study was to introduce a simultaneous approach to extracting growth trends from diameter increment series by combining tree- and stand-level information in the framework of mixed-effects modeling. The model results are compared with those of the sequential modeling approach according to Yue et al. Can J For Res 41:1577–1589, (2011). Comprehensive data from periodically repeated tree and stand measurements on 113 Norway spruce long-term forest research stands distributed throughout southwestern Germany as well as annually resolved diameter growth data from stem analysis of 581 sample trees are used for model parameterization. Results provide clear evidence for distinct temporal variation due to environment-related effects after accounting for tree- and stand-level effects on diameter growth of Norway spruce over the past 136 years, with especially high growth levels during the last four decades. Most remarkable is the considerable increase in diameter growth rates following the 1947–1952 growth depression. From the early 1960s onwards, growth rates fluctuated on an elevated level until most recent years. Model comparison reveals that results of the sequential and simultaneous modeling approaches are similar with respect to the course of the time-specific environmental effects on growth. We conclude that the proposed simultaneous modeling approach has the advantages of extracting growth trends at a higher level of precision and being the more parsimonious modeling option.

Predicting estuarine benthic production using functional diversity

Event Abstract Back to Event Predicting estuarine benthic production using functional diversity Marina Dolbeth1, 2*, Sylvain Dolédec3 and Miguel Â. Pardal2 1 University of Aveiro, Dep. Biology, Portugal 2 University of Coimbra, Portugal 3 Université Lyon 1, France We considered an estuarine system having naturally low levels of diversity, but attaining considerable high production levels, and being subjected to different sorts of anthropogenic impacts and climate events to investigate the relationship between diversity and secondary production. Functional diversity measures were used to predict benthic production, which is considered as a proxy of the ecosystem provisioning services. To this end, we used a 14-year dataset on benthic invertebrate community production from a seagrass and a sandflat habitat and we adopted a sequential modeling approach, where abiotic, trait community weighted means (CWM) and functional diversity indices were tested by generalized linear models (GLM), and their significant variables were then combined to produce a final model. Almost 90% of variance of the benthic production could be predicted by combining the number of locomotion types, the absolute maximum atmospheric temperature (proxy of the heat waves occurrence), the type of habitat and the mean body mass, by order of importance. This result is in agreement with the mass ratio hypothesis, where ecosystem functions/services can be chiefly predicted by the dominant trait in the community, here measured as CWM. The increase of benthic production with the number of locomotion types may be seen as greater possibility of using the resources available in the system. Such greater efficiency would increase production. The other variables were also discussed in line of the previous hypothesis and taking into account the general positive relationship obtained between production and functional diversity indices. Overall, it was concluded that traits representative of wider possibilities of using available resources and higher functional diversity are related with higher benthic production. Acknowledgements This research was supported by FCT (Portuguese Foundation for Science and Technology), through a grant attributed to M Dolbeth (SFRH/BPD/41117/2007) and BIOCHANGED project (PTDC/MAR/111901/2009), subsidized by the European Social Fund and MCTES (Portuguese Ministry of Science, Technology and Higher Education), through the POPH (Human Potential Operational Programme), QREN (National Strategic Reference Framework) and COMPETE (Programa Operacional Factores de Competitividade). Keywords: Secondary Production, functional diversity, benthic communities, ecosystem services, Estuaries Conference: IMMR | International Meeting on Marine Research 2014, Peniche, Portugal, 10 Jul - 11 Jul, 2014. Presentation Type: Oral Presentation Topic: BIODIVERSITY, CONSERVATION AND COASTAL MANAGEMENT Citation: Dolbeth M, Dolédec S and Pardal M (2014). Predicting estuarine benthic production using functional diversity. Front. Mar. Sci. Conference Abstract: IMMR | International Meeting on Marine Research 2014. doi: 10.3389/conf.fmars.2014.02.00092 Copyright: The abstracts in this collection have not been subject to any Frontiers peer review or checks, and are not endorsed by Frontiers. They are made available through the Frontiers publishing platform as a service to conference organizers and presenters. The copyright in the individual abstracts is owned by the author of each abstract or his/her employer unless otherwise stated. Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed. For Frontiers’ terms and conditions please see https://www.frontiersin.org/legal/terms-and-conditions. Received: 10 May 2014; Published Online: 18 Jul 2014. * Correspondence: Dr. Marina Dolbeth, University of Aveiro, Dep. Biology, Aveiro, Portugal, marinadolbeth@gmail.com Login Required This action requires you to be registered with Frontiers and logged in. To register or login click here. Abstract Info Abstract The Authors in Frontiers Marina Dolbeth Sylvain Dolédec Miguel  Pardal Google Marina Dolbeth Sylvain Dolédec Miguel  Pardal Google Scholar Marina Dolbeth Sylvain Dolédec Miguel  Pardal PubMed Marina Dolbeth Sylvain Dolédec Miguel  Pardal Related Article in Frontiers Google Scholar PubMed Abstract Close Back to top Javascript is disabled. Please enable Javascript in your browser settings in order to see all the content on this page.

Internet shopping and internet banking in sequence: An explanatory case study

Adoption of Internet banking often follows on from usage of Internet shopping, but policies to increase Internet banking use typically ignore this ordering. This article presents a case study that underscores this sequence of Internet service adoption and identifies factors that shape the propensity to use the Internet for shopping and banking. Application of bivariate probit regression techniques to data sourced from a survey of 259 respondents in Athens, Greece, and estimation of marginal effects of the determinants of Internet banking use conditioned on the determinants of Internet shopping use illustrate that ignoring the sequence of Internet service use can lead to incorrect policy recommendations. This article contributes to the literature by theorising the underlying causal mechanisms of Internet banking adoption and presenting supporting evidence via a sequential modelling approach. We find that personal capacity is an important determinant of Internet banking use in a standard, non-sequential approach but it has no significant effect when the model is sequential. Our results suggest that policymakers should emphasise usefulness attributes of Internet banking when attempting to increase Internet banking usage by people who already use the Internet for shopping.

Modeling the free vibrations of an acoustic guitar top plate

Using computer models to simulate the sound of an acoustic guitar can significantly decrease lead time for new designs. In order to create an accurate computer model of a Dreadnought-style acoustic guitar, a sequential modeling approach was used. A finite element model of a bare top plate with braces and a bridge plate was created. The top and plate and braces were modeled as plate elements with orthotropic material properties. The natural variation of the wood properties was also examined along with their dependence on moisture content. The modes of the model were then compared to experimentally obtained modes from top plate prototypes. The modeshapes of the model compared well to those measured. Uncertainty analysis was also performed and the statistical bound of natural error between wood samples was determined to be approximately 8%. The natural frequencies of the model fell within the error bound for lower-order modes but diverged slightly for several higher-order modes. These results indicate the importance of using accurate material properties in models of acoustic guitars.

Application of Hematological Toxicity Modeling in Clinical Development of Abexinostat (S-78454, PCI-24781), A New Histone Deacetylase Inhibitor

A population pharmacokinetic/pharmacodynamic (PK/PD) model was developed to describe the thrombocytopenia (dose-limiting toxicity) of abexinostat, a new histone deacetylase inhibitor. An optimal administration schedule of the drug was determined using a simulation-based approach. Early PK and PK/PD data were analysed using a sequential population modeling approach (NONMEM 7), allowing for the description of a PK profile and platelet-count decrease after abexinostat administration with various administration schedules. Simulations of platelet count with several administration schedules over 3-week treatment cycles (ASC) and over a day (ASD) were computed to define the optimal schedule that limits the depth of thrombocytopenia. An intermediate PK/PD model accurately described the data. The administration of abexinostat during the first 4 days of each week in a 3-week cycle resulted in fewer adverse events (with no influence of ASD on platelet count profiles), and corresponded to the optimal treatment schedule. This administration schedule was clinically evaluated in a phase I clinical trial and allowed for the definition of a new maximum tolerated dose (MTD), leading to a nearly 30% higher dose-intensity than that of another previously tested schedule. Lastly, a final model was built using all of the available data. The final model, characterizing the dose-effect and the dose-toxicity relationships, provides a useful modeling tool for clinical drug development.

Modeling Parking Behavior Under Uncertainty: A Static Game Theoretic versus a Sequential Neo-additive Capacity Modeling Approach

This paper proposes two types of parking choice models, a static game theoretic model and a dynamic neo-additive capacity model, to capture the competition among drivers for limited desirable parking spaces. The static game assumes that drives make decisions simultaneously and with perfect knowledge about the characteristics of the parking system and the strategies of their fellow drivers in the system; the model thus captures only the rational aspect of parking choice behavior and pays no attention to modeling individual drivers’ psychological characteristics. The dynamic model, on the other hand, considers individual drivers’ psychological characteristics under uncertainty (i.e. optimistic and pessimistic attitudes) and thus captures the impacts of the irrational side of parking behavior in addition to the rational aspect. Following the formulation of the two models, they are both used to predict parking behavior as observed on a set of parking lots on the University at Buffalo north campus. Specifically for the dynamic model, the model is first calibrated based on real data collected from video recorded observations for a pair of parking lots, and then used to predict behavior on another pair. Validation results show higher predictive accuracy for the dynamic neo-additive capacity model compared to the static game theoretic model. This in turn suggests that the psychological characteristics of drivers play an important role in the parking lot choice decision process, and points to the potential for parking information systems to eliminate the unnecessary additional traffic generated by the parking search process.

Population PK/PD analysis of metformin using the signal transduction model

Metformin, a biguanide glucose lowering agent, is commonly used to manage type 2 diabetes. The molecular mechanisms of metformin have not been fully identified, but turnover of biomarkers such as glucose and signalling pathways or translocation of glucose transporters are closely related to the glucose-lowering effects of metformin. The PK/PD of metformin have been investigated in healthy humans and patients with type 2 diabetes mellitus and modelling has been performed using an indirect response model. The purpose of this investigation was to develop a population PK/PD model for metformin using a signal transduction model in healthy humans and predict the PK/PD profile in patients with type 2 diabetes. The aim was to compare a previous model (a biophase model) with the signal transduction model, and use a more appropriate model to follow the actions of metformin. Additionally, our developed model was appropriate to predict the time course of plasma metformin and fasting plasma glucose (FPG) concentrations in patients with type 2 diabetes. To our knowledge, this is the first published population PK/PD analysis using the signal transduction model for metformin. AIMS To develop a population pharmacokinetic (PK) and pharmacodynamic (PD) model for metformin (500 mg) using the signal transduction model in healthy humans and to predict the PK/PD profile in patients with type 2 diabetes. Following the oral administration of 500 mg metformin to healthy humans, plasma concentrations of metformin were measured using LC-MS/MS. A sequential modelling approach using NONMEM VI was used to facilitate data analysis. Monte Carlo simulation was performed to predict the antihyperglycaemic effect in patients with type 2 diabetes. Forty-two healthy humans were included in the study. Population mean estimates (relative standard error, RSE) of apparent clearance, apparent volume of distribution and the absorption rate constant were 52.6 l h(-1) (4.18%), 113 l (56.6%) and 0.41 h(-1) , respectively. Covariate analyses revealed that creatinine clearance (CL(CR) ) significantly influenced metformin: CL/F= 52.6 × (CL(cr) /106.5)(0.782) . The signal transduction model was applied to describe the antihyperglycaemic effect of metformin. The population means for efficacy, potency, transit time and the Hill coefficient were estimated to be 19.8 (3.17%), 3.68 µg ml(-1) (3.89%), 0.5 h (2.89%) and 0.547 (9.05%), respectively. The developed model was used to predict the antihyperglycaemic effect in patients with type 2 diabetes. The predicted plasma glucose concentration value was similar to previous values. The population signal transduction model was developed and evaluated for metformin use in healthy volunteers. Model evaluation by non-parametric bootstrap analysis suggested that the proposed model was robust and parameter values were estimated with good precision.

Method of scale bridging for thermoelasticity of cross-linked epoxy/SiC nanocomposites at a wide range of temperatures

In this study, we propose a sequential multiscale modeling approach to describe the thermoelastic behavior of cross-linked epoxy/silicon carbide (SiC) nanocomposites embedding different radii of nanofiller through molecular dynamics (MD) simulations and a continuum micromechanics constitutive model. In MD simulations, the coefficients of thermal expansion (CTEs), the elastic moduli, and the glass transition temperatures of different nanocomposites having different particle sizes are obtained at temperatures from 250 K to 550 K, the range in which cross-linked epoxy polymers generally experience the glassy-to-rubbery phase transition and consequently, their CTEs and elastic moduli change dramatically. In the equivalent continuum model, an interphase is defined between the particle and the matrix to account for the contribution of the polymer densification in the vicinity of the nanoparticle to the size-dependent CTE and elastic modulus at each temperature. Based on the thermal strain field defined in the micromechanics constitutive model, a physically meaningful description of the thermal expansion behavior of the interphase is obtained to reproduce the MD simulation results from fully continuum-based approaches for nanocomposites in rubbery state as well as in glassy state. Finally, the accuracy of the proposed multiscale approach is confirmed from finite element analysis.

Characterization of Compressive Die Stresses in CBGA Microprocessor Packaging Due to Component Assembly and Heat Sink Clamping

Microprocessor packaging in modern workstations and servers often consists of one or more large flip chip die that are mounted to a high performance ceramic chip carrier. The final assembly configuration features a complex stack up of flip chip area array solder interconnects, underfill, ceramic substrate, lid, heat sink, thermal interface materials (TIMs), second level ceramic ball grid array (CBGA) solder joints, organic printed circuit board, etc., so that a very complicated set of loads is transmitted to the microprocessor chip. Several trends in the evolution of this packaging architecture have exacerbated die stress levels including the transition to larger die, high coefficient of thermal expansion (CTE) ceramic substrates, lead free solder joints, higher levels of power generation, and larger heat sinks with increased clamping forces. Die stress effects are of concern due to several reasons including degradation of silicon device performance (mobility/speed), damage that can occur to the copper/low-k top level interconnect layers, and potential mechanical failure of the silicon in extreme cases. In this work, test chips containing piezoresistive stress sensors have been used to measure the buildup of mechanical stresses in a microprocessor die after various steps of the flip chip CBGA assembly process. The utilized (111) silicon test chips were able to measure the complete three-dimensional stress state at each sensor site being monitored by the data acquisition hardware. Special test fixtures were developed to eliminate any additional stresses due to clamping effects. The developed normal stresses are compressive (triaxial compression) across the die surface, with significant in-plane and out-of-plane (interfacial) shear stresses also present at the die corners. The compressive stresses increase with each assembly step (flip chip solder joint reflow, underfill dispense and cure, and lid attachment). The experimental observations from this study show clearly that large area array flip chips are subjected to relatively large compressive in-plane normal stresses after solder reflow. We also observed that the majority of the die compressive stress is accumulated during the underfilling assembly step. Typical increases in the stress magnitude were on the order of 300% (relative to the stresses due to solder joint reflow only). As a general “rule of thumb,” approximately two-thirds (∼66%) of the final die stress magnitudes were observed to be developed during the underfill dispense and cure, with the second largest contribution coming from the die attachment, and the smallest contribution coming from lid attachment. The experimental test chip stress measurements were correlated with finite element simulations of the packaging process. A sequential modeling approach has been utilized to predict the build-up of compressive stress. The utilized method incorporates precise thermal histories of the packaging process, element creation, and nonlinear temperature and time dependent material properties. With suitable detail in the models, excellent correlation has been obtained with the sensor data throughout all packaging processes. Finally, CBGAs with the stress sensing chips were soldered to organic printed circuit board (PCB) test boards. A simulated heat sink loading applied, and the stresses were measured as a function of the clamping force. Compressive stress changes of up to − 60 MPa were observed for a 1000 N applied clamping force. The experimental test chip stress measurements were correlated with finite element simulations of the clamping process. With suitable detail in the models, excellent correlation has been obtained for the stress changes occurring during simulated heat sink clamping.

Predictive model for carbon nanotube–reinforced nanocomposite modulus driven by micromechanical modeling and physical experiments

This article proposes an improved surrogate model for the prediction of the elastic modulus of carbon nanotube–reinforced-nanocomposites. By statistically combining micromechanical modeling results with limited amounts of experimental data, a better predictive surrogate model is constructed using a two-stage sequential modeling approach. A set of data for multi-walled carbon nanotube–bismaleimide nanocomposites is used in a case study to demonstrate the effectiveness of the proposed surrogate modeling procedure. In the case study, the theoretical composite modulus is computed with micromechanical models, and the experimental modulus is measured through tensile tests. Both theoretical and experimental composite moduli are integrated by using a statistical adjustment method to construct the surrogate model. The results demonstrate an improved predictive ability compared to the original micromechanical model.

Inferring Upon Heterogeneous Associations in Dairy Cattle Performance Using a Bivariate Hierarchical Model

Multivariate hierarchical Bayesian models provide a flexible framework for comprehensive study of biological systems with more than one outcome. Recent methodological developments facilitate modeling of heterogeneous associations between outcomes by specifying a linear mixed model on (co)variances at different levels of the data structure. Motivated by previous evidence for heterogeneous correlations in animal agriculture, we apply the proposed hierarchical Bayesian models to study the nature of the correlations between key performance outcomes in dairy cattle production systems, namely milk yield and reproduction. That is, the association between these outcomes might depend upon various fixed and random effect sources of heterogeneity both at the individual cow (residual) level as well as the herd (cluster) level. We thus propose a sequential modeling approach based on the deviance information criterion to select relevant explanatory variables on both types of associations. Furthermore, we extend the proposed methodology to accommodate right-censored outcomes, as common for dairy reproduction data, and use it to analyze field data from the Michigan dairy industry. The nature of the associations between milk production and reproduction in dairy cattle was inferred to be strongly heterogeneous and driven by multiple farm management practices and herd attributes, as well as by random clustering effects, at both cow and herd levels, thereby suggesting potential between-herd and within-herd intervention strategies to optimize performance of dairy production systems. Supplementary materials are available online.

Exploring a Method of Estimating Vehicle Miles of Travel

This research attempts to develop a statistical model that can analyze the impact of land use on vehicle use. By analyzing travel behavior data from 2009 National Household Travel Survey (NHTS) data, a sequential modeling approach with five sub?models, called the SCAG NHTS Model, was developed to estimate the land use-travel behavior relationship, particularly for household Vehicle Miles of Travel (VMT). The SCAG NHTS Model was developed using the parcel database on socioeconomic and land use data. The test for model validity and sensitivity has shown that the SCAG NHTS Model produces reasonable and accountable results and is sensitive to land use characteristics. The SCAG NHTS Model also produced model results similar to SCAG’s trip?based model. After running the model with SCAG 2012 Regional Transportation Plan scenario data, the SCAG NHTS Model estimates household VMT, which is 2.4 % lower than the trip?based model.

A sequential modelling approach to assess groundwater–surface water resources in a snow dominated region of Finland

Methodology was developed to estimate surface water–groundwater interactions in a changing climate in cold, snow-dominated regions and tested on an unconfined esker aquifer in northern Finland. The Watershed Simulation and Forecasting System (WSFS) model was used to simulate surface water levels. A coupled heat and mass transfer model for soil–plant–atmosphere systems (CoupModel) was used to simulate recharge. The CoupModel is able to simulate water flow through frozen soil. The simulated surface water level and recharge data were linked to the Groundwater Modelling System (GMS) version 6.5, and the three-dimensional groundwater flow model MODFLOW was used to study the impact of climate variability on groundwater levels and groundwater–surface water interactions. In addition, under A1B climate change scenario changes in aquifer storage and groundwater–surface water interactions were studied through the end of the 21st century. Recharge was shown to be sensitive to changes in soil frost and hence modelling approaches should include a soil frost component to account for the impact of frost on hydraulic conductivity. Under A1B climate change scenario study, groundwater recharge is projected to increase in winter months due to increase in snowmelt and decrease in soil frost depth. The spring snowmelt peak in late spring will decrease. This will reduce aquifer storage in early spring, increasing the vulnerability to summer droughts.

Population pharmacokinetics and pharmacodynamics of hydroxyurea in sickle cell anemia patients, a basis for optimizing the dosing regimen

BackgroundHydroxyurea (HU) is the first approved pharmacological treatment of sickle cell anemia (SCA). The objectives of this study were to develop population pharmacokinetic(PK)-pharmacodynamic(PD) models for HU in order to characterize the exposure-efficacy relationships and their variability, compare two dosing regimens by simulations and develop some recommendations for monitoring the treatment.MethodsThe models were built using population modelling software NONMEM VII based on data from two clinical studies of SCA adult patients receiving 500-2000 mg of HU once daily. Fetal hemoglobin percentage (HbF%) and mean corpuscular volume (MCV) were used as biomarkers for response. A sequential modelling approach was applied. Models were evaluated using simulation-based techniques. Comparisons of two dosing regimens were performed by simulating 10000 patients in each arm during 12 months.ResultsThe PK profiles were described by a bicompartmental model. The median (and interindividual coefficient of variation (CV)) of clearance was 11.6 L/h (30%), the central volume was 45.3 L (35%). PK steady-state was reached in about 35 days. For a given dosing regimen, HU exposure varied approximately fivefold among patients. The dynamics of HbF% and MCV were described by turnover models with inhibition of elimination of response. In the studied range of drug exposures, the effect of HU on HbF% was at its maximum (median Imax was 0.57, CV was 27%); the effect on MCV was close to its maximum, with median value of 0.14 and CV of 49%. Simulations showed that 95% of the steady-state levels of HbF% and MCV need 26 months and 3 months to be reached, respectively. The CV of the steady-state value of HbF% was about 7 times larger than that of MCV. Simulations with two different dosing regimens showed that continuous dosing led to a stronger HbF% increase in some patients.ConclusionsThe high variability of response to HU was related in part to pharmacokinetics and to pharmacodynamics. The steady-state value of MCV at month 3 is not predictive of the HbF% value at month 26. Hence, HbF% level may be a better biomarker for monitoring HU treatment. Continuous dosing might be more advantageous in terms of HbF% for patients who have a strong response to HU.Trial RegistrationThe clinical studies whose data are analysed and reported in this work were not required to be registered in France at their time. Both studies were approved by local ethics committees (of Mondor Hospital and of Kremlin-Bicetre Hospital) and written informed consent was obtained from each patient.

Sequential modelling of a full-scale wastewater treatment plant using an artificial neural network

This work proposes a sequential modelling approach using an artificial neural network (ANN) to develop four independent multivariate models that are able to predict the dynamics of biochemical oxygen demand (BOD), chemical oxygen demand (COD), suspended solid (SS), and total nitrogen (TN) removal in a wastewater treatment plant (WWTP). Suitable structures of ANN models were automatically and conveniently optimized by a genetic algorithm rather than the conventional trial and error method. The sequential modelling approach, which is composed of two parts, a process disturbance estimator and a process behaviour predictor, was also presented to develop multivariate dynamic models. In particular, the process disturbance estimator was first employed to estimate the influent quality. The process behaviour predictor then sequentially predicted the effluent quality based on the estimated influent quality from the process disturbance estimator with other process variables. The efficiencies of the developed ANN models with a sequential modelling approach were demonstrated with a practical application using a data set collected from a full-scale WWTP during 2years. The results show that the ANN with the sequential modelling approach successfully developed multivariate dynamic models of BOD, COD, SS, and TN removal with satisfactory estimation and prediction capability. Thus, the proposed method could be used as a powerful tool for the prediction of complex and nonlinear WWTP performance.

An adaptation of a prototype recursive model for evaluating and predicting micro-macro economic data in Botswana

Naturally arising patterns in data are known to present potential sources of useful information at both micro and macro-economic levels. We carry out unsupervised and supervised modelling of Botswana's macro-economic data attributes obtained from disparate sources. Both techniques, commonly used to detect inherent patterns in data, have adjustable parameters which inevitably vary across applications. Thus, we propose a sequential unsupervised-supervised modelling approach in which Exploratory Data Analysis (EDA) is used to detect basic structures in data which are then passed on an algorithm based on the Expectation-Maximisation (EM) mechanics. The EM convergent values are then used to guide data labelling before applying the neural networks model. We demonstrate how future economic structures may be detected, monitored and managed by iteratively focusing on conditional checks of a generic algorithm. For the purposes of modelling robustness, we propose setting up an integrated data repository and source that would provide data-based guidelines to policy makers in addressing the country's economic issues while providing economic researchers access to data and/or information resources. Outstanding issues are identified and discussed and potential future directions are clearly highlighted. Key words: Data mining, data over-fitting, data recycling, EM algorithm, neural networks, knowledge extraction from data, supervised learning, unsupervised learning.

The use of the SAEM algorithm in MONOLIX software for estimation of population pharmacokinetic-pharmacodynamic-viral dynamics parameters of maraviroc in asymptomatic HIV subjects

Using simulated viral load data for a given maraviroc monotherapy study design, the feasibility of different algorithms to perform parameter estimation for a pharmacokinetic-pharmacodynamic-viral dynamics (PKPD-VD) model was assessed. The assessed algorithms are the first-order conditional estimation method with interaction (FOCEI) implemented in NONMEM VI and the SAEM algorithm implemented in MONOLIX version 2.4. Simulated data were also used to test if an effect compartment and/or a lag time could be distinguished to describe an observed delay in onset of viral inhibition using SAEM. The preferred model was then used to describe the observed maraviroc monotherapy plasma concentration and viral load data using SAEM. In this last step, three modelling approaches were compared; (i) sequential PKPD-VD with fixed individual Empirical Bayesian Estimates (EBE) for PK, (ii) sequential PKPD-VD with fixed population PK parameters and including concentrations, and (iii) simultaneous PKPD-VD. Using FOCEI, many convergence problems (56%) were experienced with fitting the sequential PKPD-VD model to the simulated data. For the sequential modelling approach, SAEM (with default settings) took less time to generate population and individual estimates including diagnostics than with FOCEI without diagnostics. For the given maraviroc monotherapy sampling design, it was difficult to separate the viral dynamics system delay from a pharmacokinetic distributional delay or delay due to receptor binding and subsequent cellular signalling. The preferred model included a viral load lag time without inter-individual variability. Parameter estimates from the SAEM analysis of observed data were comparable among the three modelling approaches. For the sequential methods, computation time is approximately 25% less when fixing individual EBE of PK parameters with omission of the concentration data compared with fixed population PK parameters and retention of concentration data in the PD-VD estimation step. Computation times were similar for the sequential method with fixed population PK parameters and the simultaneous PKPD-VD modelling approach. The current analysis demonstrated that the SAEM algorithm in MONOLIX is useful for fitting complex mechanistic models requiring multiple differential equations. The SAEM algorithm allowed simultaneous estimation of PKPD and viral dynamics parameters, as well as investigation of different model sub-components during the model building process. This was not possible with the FOCEI method (NONMEM version VI or below). SAEM provides a more feasible alternative to FOCEI when facing lengthy computation times and convergence problems with complex models.