Search

International audience; We have performed a parallel tempering crankshaft motion Monte Carlo simulation on a model of the GABA type A receptor with the aim of exploring a wide variety of local conformational space. We develop a novel method to analyse the protein movements in terms of a correlation tensor and use this to explore the gating process, that is, how agonist binding could cause ion channel opening. We find that simulated binding impulses to varying clusters of GABA binding site residues produce channel opening, and that equivalent impulses to single GABA sites produce partial opening.

International audience; A model and data toolbox is presented to assess risks from combined exposure to multiple chemicals using probabilistic methods. The Monte Carlo Risk Assessment (MCRA) toolbox, also known as the EuroMix toolbox, has more than 40 modules addressing all areas of risk assessment, and includes a data repository with data collected in the EuroMix project. This paper gives an introduction to the toolbox and illustrates its use with examples from the EuroMix project. The toolbox can be used for hazard identification, hazard characterisation, exposure assessment and risk characterisation. Examples for hazard identification are selection of substances relevant for a specific adverse outcome based on adverse outcome pathways and QSAR models. Examples for hazard characterisation are calculation of benchmark doses and relative potency factors with uncertainty from dose response data, and use of kinetic models to perform in vitro to in vivo extrapolation. Examples for exposure assessment are assessing cumulative exposure at external or internal level, where the latter option is needed when dietary and non-dietary routes have to be aggregated. Finally, risk characterisation is illustrated by calculation and display of the margin of exposure for single substances and for the cumulation, including uncertainties derived from exposure and hazard characterisation estimates.

International audience; Proton imaging can be carried out on microscopic samples by focusing the beam to a diameter ranging from a few micrometers down to a few tens of nanometers, depending on the required beam intensity and spatial resolution. Three-dimensional (3D) imaging by tomography is obtained from proton transmission (STIM: Scanning Transmission Ion Microscopy) and/or X-ray emission (PIXE: Particle Induced X-ray Emission). In these experiments, the samples are dehydrated for under vacuum analysis. In situ quantification of nanoparticles has been carried out at CENBG in the frame of nanotoxicology studies, on cells and small organisms used as biological models, especially on Caenorhabditis elegans (C. elegans) nematodes. Tomography experiments reveal the distribution of mass density and chemical content (in g.cm−3) within the analyzed volume. These density values are obtained using an inversion algorithm. To investigate the effect of this data reduction process, we defined different numerical phantoms, including a (dehydrated) C. elegans phantom whose geometry and density were derived from experimental data. A Monte Carlo simulation based on the Geant4 toolkit was developed. Using different simulation and reconstruction conditions, we compared the resulting tomographic images to the initial numerical reference phantom. A study of the relative error between the reconstructed and the reference images lead to the result that 20 protons per shot can be considered as an optimal number for 3D STIM imaging. Preliminary results for PIXE tomography are also presented, showing the interest of such numerical phantoms to produce reference data for future studies on X-ray signal attenuation in thick samples.

Dengue dynamics are shaped by the complex interplay between several factors, including vector seasonality, interaction between four virus serotypes, and inapparent infections. However, paucity or quality of data do not allow for all of these to be taken into account in mathematical models. In order to explore separately the importance of these factors in models, we combined surveillance data with a local-scale cluster study in the rural province of Kampong Cham (Cambodia), in which serotypes and asymptomatic infections were documented. We formulate several mechanistic models, each one relying on a different set of hypotheses, such as explicit vector dynamics, transmission via asymptomatic infections and coexistence of several virus serotypes. Models are confronted with the observed time series using Bayesian inference, through Markov chain Monte Carlo. Model selection is then performed using statistical information criteria, and the coherence of epidemiological characteristics (reproduction numbers, incidence proportion, dynamics of the susceptible classes) is assessed in each model. Our analyses on transmission dynamics in a rural endemic setting highlight that two-strain models with interacting effects better reproduce the long term data, but they are difficult to parameterize when relying on incidence cases only. On the other hand, considering the available data, incorporating vector and asymptomatic components seems of limited added-value when seasonality and underreporting are already accounted for. Keywords: Model selection, Dengue, Vector borne disease, Mathematical model, Cambodia

International audience; Significant alterations in the pharmacokinetics (PK) of antimicrobials have been reported in critically ill patients. We describe PK parameters of imipenem in intensive care unit (ICU) patients with suspected ventilator-associated pneumonia and evaluate several dosage regimens. This French multicentre, prospective, open-label study was conducted in ICU patients with a presumptive diagnosis of ventilator-associated pneumonia caused by Gram-negative bacilli, who empirically received imipenem intravenously every 8 h. Plasma imipenem concentrations were measured during the fourth imipenem infusion using six samples (trough, 0.5, 1, 2, 5 and 8 h). Data were analysed with a population approach using the stochastic approximation expectation maximization algorithm in Monolix 4.2. A Monte Carlo simulation was performed to evaluate the following six dosage regimens: 500, 750 or 1000 mg with administration every 6 or 8 h. The pharmacodynamic target was defined as the probability of achieving a fractional time above the minimal inhibitory concentration (MIC) of >40%. Fifty-one patients were included in the PK analysis. Imipenem concentration data were best described by a two-compartment model with three covariates (creatinine clearance, total bodyweight and serum albumin). Estimated clearance (between-subject variability) was 13.2 l h(-1) (38%) and estimated central volume 20.4 l (31%). At an MIC of 4 μg ml(-1) , the probability of achieving 40% fractional time > MIC was 91.8% for 0.5 h infusions of 750 mg every 6 h, 86.0% for 1000 mg every 8 h and 96.9% for 1000 mg every 6 h. This population PK model accurately estimated imipenem concentrations in ICU patients. The simulation showed that for these patients, the best dosage regimen of imipenem is 750 mg every 6 h and not 1000 mg every 8 h.

International audience; The activity of a neural network is defined by patterns of spiking and silence from the individual neurons. Because spikes are (relatively) sparse, patterns of activity with increasing numbers of spikes are less probable, but, with more spikes, the number of possible patterns increases. This tradeoff between probability and numerosity is mathematically equivalent to the relationship between entropy and energy in statistical physics. We construct this relationship for populations of up to N = 160 neurons in a small patch of the vertebrate retina, using a combination of direct and model-based analyses of experiments on the response of this network to naturalistic movies. We see signs of a thermodynamic limit, where the entropy per neuron approaches a smooth function of the energy per neuron as N increases. The form of this function corresponds to the distribution of activity being poised near an unusual kind of critical point. We suggest further tests of criticality, and give a brief discussion of its functional significance.

International audience; Significant alterations in the pharmacokinetics (PK) of antimicrobials have been reported in critically ill patients. We describe PK parameters of imipenem in intensive care unit (ICU) patients with suspected ventilator-associated pneumonia and evaluate several dosage regimens. This French multicentre, prospective, open-label study was conducted in ICU patients with a presumptive diagnosis of ventilator-associated pneumonia caused by Gram-negative bacilli, who empirically received imipenem intravenously every 8 h. Plasma imipenem concentrations were measured during the fourth imipenem infusion using six samples (trough, 0.5, 1, 2, 5 and 8 h). Data were analysed with a population approach using the stochastic approximation expectation maximization algorithm in Monolix 4.2. A Monte Carlo simulation was performed to evaluate the following six dosage regimens: 500, 750 or 1000 mg with administration every 6 or 8 h. The pharmacodynamic target was defined as the probability of achieving a fractional time above the minimal inhibitory concentration (MIC) of >40%. Fifty-one patients were included in the PK analysis. Imipenem concentration data were best described by a two-compartment model with three covariates (creatinine clearance, total bodyweight and serum albumin). Estimated clearance (between-subject variability) was 13.2 l h(-1) (38%) and estimated central volume 20.4 l (31%). At an MIC of 4 μg ml(-1) , the probability of achieving 40% fractional time > MIC was 91.8% for 0.5 h infusions of 750 mg every 6 h, 86.0% for 1000 mg every 8 h and 96.9% for 1000 mg every 6 h. This population PK model accurately estimated imipenem concentrations in ICU patients. The simulation showed that for these patients, the best dosage regimen of imipenem is 750 mg every 6 h and not 1000 mg every 8 h.

International audience; Significant alterations in the pharmacokinetics (PK) of antimicrobials have been reported in critically ill patients. We describe PK parameters of imipenem in intensive care unit (ICU) patients with suspected ventilator-associated pneumonia and evaluate several dosage regimens. This French multicentre, prospective, open-label study was conducted in ICU patients with a presumptive diagnosis of ventilator-associated pneumonia caused by Gram-negative bacilli, who empirically received imipenem intravenously every 8 h. Plasma imipenem concentrations were measured during the fourth imipenem infusion using six samples (trough, 0.5, 1, 2, 5 and 8 h). Data were analysed with a population approach using the stochastic approximation expectation maximization algorithm in Monolix 4.2. A Monte Carlo simulation was performed to evaluate the following six dosage regimens: 500, 750 or 1000 mg with administration every 6 or 8 h. The pharmacodynamic target was defined as the probability of achieving a fractional time above the minimal inhibitory concentration (MIC) of >40%. Fifty-one patients were included in the PK analysis. Imipenem concentration data were best described by a two-compartment model with three covariates (creatinine clearance, total bodyweight and serum albumin). Estimated clearance (between-subject variability) was 13.2 l h(-1) (38%) and estimated central volume 20.4 l (31%). At an MIC of 4 μg ml(-1) , the probability of achieving 40% fractional time > MIC was 91.8% for 0.5 h infusions of 750 mg every 6 h, 86.0% for 1000 mg every 8 h and 96.9% for 1000 mg every 6 h. This population PK model accurately estimated imipenem concentrations in ICU patients. The simulation showed that for these patients, the best dosage regimen of imipenem is 750 mg every 6 h and not 1000 mg every 8 h.

When information is sparse, individual parameters derived from a non-linear mixed effects model analysis can shrink to the mean. The objective of this work was to predict individual parameter shrinkage from the Bayesian information matrix (M BF ). We 1) Propose and evaluate an approximation of M BF by First-Order linearization (FO), 2) Explore by simulations the relationship between shrinkage and precision of estimates and 3) Evaluate prediction of shrinkage and individual parameter precision. We approximated M BF using FO. From the shrinkage formula in linear mixed effects models, we derived the predicted shrinkage from M BF . Shrinkage values were generated for parameters of two pharmacokinetic models by varying the structure and the magnitude of the random effect and residual error models as well as the design. We then evaluated the approximation of M BF FO and compared it to Monte-Carlo (MC) simulations. We finally compared expected and observed shrinkage as well as the predicted and estimated Standard Errors (SE) of individual parameters. M BF FO was similar to M BF MC. Predicted and observed shrinkages were close . Predicted and estimated SE were similar. M BF FO enables prediction of shrinkage and SE of individual parameters. It can be used for design optimization.

International audience; Errors in the flip angle have to be corrected in many magnetic resonance imaging applications, especially for T1 quantification. However, the existing methods of B1 mapping fail to measure lower values of the flip angle despite the fact that these are extensively used in dynamic acquisition and 3D imaging. In this study, the nonlinearity of the radiofrequency (RF) transmit chain, especially for very low flip angles, is investigated and a simple method is proposed to accurately determine both the gain of the RF transmitter and the B1 field map for low flip angles. The method makes use of the spoiled gradient echo sequence with long repetition time (TR), such as applied in the double-angle method. It uses an image acquired with a flip angle of 90° as a reference image that is robust to B1 inhomogeneity. The ratio of the image at flip angle alpha to the image at a flip angle of 90° enables us to calculate the actual value of alpha. This study was carried out at 1.5 and 4.7 T, showing that the linearity of the RF supply system is highly dependent on the hardware. The method proposed here allows us to measure the flip angle from 1° to 60° with a maximal uncertainty of 10% and to correct T1 maps based on the variable flip angle method.

International audience; We address the issue of jointly detecting brain activity and estimating underlying brain hemodynamics from functional MRI data. We adopt the so-called Joint Detection Estimation (JDE) framework that takes spatial dependencies between voxels into account. We recast the JDE into a missing data framework and derive a Variational Expectation-Maximization (VEM) algorithm for its inference. It follows a new algorithm that has interesting advantages over the previously used intensive simulation methods (Markov Chain Monte Carlo, MCMC): tests on artificial data show that the VEM-JDE is more robust to model mis-specification while additional tests on real data confirm that it achieves similar performance in much less computation time.

International audience; The decomposition of multiunit signals consists of the restoration of spike trains and action potentials in neural or muscular recordings. Because of the complexity of automatic decomposition, semiautomatic procedures are sometimes chosen. The main difficulty in automatic decomposition is the resolution of temporally overlapped potentials. In a previous study , we proposed a Bayesian model coupled with a maximum a posteriori (MAP) estimator for fully automatic decomposition of multiunit recordings and we showed applications to intramuscular EMG signals. In this study, we propose a more complex signal model that includes the variability in amplitude of each unit potential. Moreover, we propose the Markov Chain Monte Carlo (MCMC) simulation and a Bayesian minimum mean square error (MMSE) estimator by averaging on samples that converge in distribution to the joint posterior law. We prove the convergence property of this approach mathematically and we test the method representatively on intramuscular multiunit recordings. The results showed that its average accuracy in spike identification is greater than 90% for intramuscular signals with up to 8 concurrently active units. In addition to intramuscular signals, the method can be applied for spike sorting of other types of multiunit recordings.

Background During the last decade, oral bisphosphonates (BP) became the most widely prescribed pharmacologic class for post-menopausal osteoporosis. However, many surveys revealed the important issue of poor persistence with those drugs resulting in a failure of treatment to reduce fracture risk sufficiently. Using a published Markov model, this study analyses the economic impact of non-persistence with bisphosphonates in the context of the introduction of generics in France. Methods Direct costs of vertebral, hip and wrist fracture were assessed and included in an existing 10-year Markov model developed to analyse consequences of non-persistence. Three alternatives of comparison were set: no treatment, real-world persistence, and ideal persistence. Simulated patients' characteristics matched those from a French observational study and the real-world adherence alternative employed persistence data from published database analysis. The risk of fracture of menopausal women and the risk reduction associated with the drugs were based on results reported in clinical trials. Incremental cost-effectiveness ratios (ICERs) were calculated first between real-world adherence and no treatment alternatives, and second between ideal and real-world persistence alternatives. The cost of non-persistence was defined as the difference between total cost of ideal and real-world persistence alternatives. Results Within fractured women population, mean costs of 10-year management of fracture were significantly different between the three alternatives with €7,239 (± €4,783), €6,711 (± €4,410) and €6,134 (± €3,945) in the no-treatment, the real-world and ideal persistence alternatives, respectively (p < 0.0001). Cost-effectiveness ratio for real-world treatment persistence compared with no-treatment alternative was found dominant and as well, alternative of ideal persistence dominated the former. Each ten percentage point of persistence gain amounted to €58 per patient, and extrapolation resulted in a global annual cost of non-persistence of over €30 million to the French health care system, with a substantial transfer from hospital to pharmacy budgets. Conclusion Within term, improving persistence with oral bisphosphonates should be economically dominant on levels currently known in real-world. Given this potential savings, ambitious adherence-enhancing interventions should be considered in osteoporotic patients.

CD8+ cytotoxic T lymphocytes (CTLs), natural killer (NK) cells, B cells and target cell limitation have all been suggested to play a role in the control of SIV and HIV-1 infection. However, previous research typically studied each population in isolation leaving the magnitude, relative importance and in vivo relevance of each effect unclear. Here we quantify the relative importance of CTLs, NK cells, B cells and target cell limitation in controlling acute SIV infection in rhesus macaques. Using three different methods, we find that the availability of target cells and CD8+ T cells are important predictors of viral load dynamics. If CTL are assumed to mediate this anti-viral effect via a lytic mechanism then we estimate that CTL killing is responsible for approximately 40% of productively infected cell death, the remaining cell death being attributable to intrinsic, immune (CD8+ T cell, NK cell, B cell) -independent mechanisms. Furthermore, we find that NK cells have little impact on the death rate of infected CD4+ cells and that their net impact is to increase viral load. We hypothesize that NK cells play a detrimental role in SIV infection, possibly by increasing T cell activation., Author Summary The role of the immune response in controlling HIV infection, or its simian counterpart SIV, has not been fully elucidated. Different studies have provided evidence of a role for different components of the immune system. Unfortunately, the different components are usually studied in isolation, making it hard to determine their relative importance. Here we study three main immune cell populations, T, B and NK cells, in parallel and found that cytotoxic T cells play the largest role in the control of SIV infection, but were not responsible for the majority of infected cell death. Interestingly, we found indications that, although NK cells contribute to the killing of infected cells, the net impact of the NK response is to increase viral load.

International audience; Subcortical-cortical interactions in the language network were investigated using dynamic causal modeling of magnetoencephalographic data recorded during auditory comprehension. Participants heard sentences that either were correct or contained violations. Sentences containing violations had syntactic or prosodic violations or both. We show that a hidden source, modeling magnetically silent deep nuclei, is required to explain the data best. This is in line with recent brain imaging studies and intracranial recordings suggesting an involvement of subcortical structures in language processing. Here, the processing of syntactic and prosodic violations elicited a global increase in the amplitude of evoked responses, both at the cortical and subcortical levels. As estimated by Bayesian model averaging, this was accompanied by various changes in cortical-cortical and subcortical-cortical connectivity. The most consistent findings in relation to violations were a decrease of reentrant inputs to Heschl's gyrus (HG) and of transcallosal lateral connections. These results suggest that in conditions where one hemisphere detects a violation, possibly via fast thalamocortical (HG) loops, the intercallosal connectivity is reduced to allow independent processing of syntax (left hemisphere) and of prosody (right hemisphere). This study is the first demonstration in cognitive neuroscience that subcortical-cortical loops can be empirically investigated using noninvasive electrophysiological recordings.

International audience; Subcortical-cortical interactions in the language network were investigated using dynamic causal modeling of magnetoencephalographic data recorded during auditory comprehension. Participants heard sentences that either were correct or contained violations. Sentences containing violations had syntactic or prosodic violations or both. We show that a hidden source, modeling magnetically silent deep nuclei, is required to explain the data best. This is in line with recent brain imaging studies and intracranial recordings suggesting an involvement of subcortical structures in language processing. Here, the processing of syntactic and prosodic violations elicited a global increase in the amplitude of evoked responses, both at the cortical and subcortical levels. As estimated by Bayesian model averaging, this was accompanied by various changes in cortical-cortical and subcortical-cortical connectivity. The most consistent findings in relation to violations were a decrease of reentrant inputs to Heschl's gyrus (HG) and of transcallosal lateral connections. These results suggest that in conditions where one hemisphere detects a violation, possibly via fast thalamocortical (HG) loops, the intercallosal connectivity is reduced to allow independent processing of syntax (left hemisphere) and of prosody (right hemisphere). This study is the first demonstration in cognitive neuroscience that subcortical-cortical loops can be empirically investigated using noninvasive electrophysiological recordings.

International audience; Characterizing the behavior and robustness of enzymatic networks with numerous variables and unknown parameter values is a major challenge in biology, especially when some enzymes have counter-intuitive properties or switch-like behavior between activation and inhibition. In this paper, we propose new methodological and tool-supported contributions, based on the intuitive formalism of temporal logic, to express in a rigorous manner arbitrarily complex dynamical properties. Our multi-step analysis allows efficient sampling of the parameter space in order to define feasible regions in which the model exhibits imposed or experimentally observed behaviors. In a first step, an algorithmic methodology involving sensitivity analysis is conducted to determine bifurcation thresholds for a limited number of model parameters or initial conditions. In a second step, this boundary detection is supplemented by a global robustness analysis, based on quasi-Monte Carlo approach that takes into account all model parameters. We apply this method to a well-documented enzymatic reaction network describing collagen proteolysis by matrix metalloproteinase MMP2 and membrane type 1 metalloproteinase (MT1-MMP) in the presence of tissue inhibitor of metalloproteinase TIMP2. For this model, our method provides an extended analysis and quantification of network robustness toward paradoxical TIMP2 switching activity between activation or inhibition of MMP2 production. Further implication of our approach is illustrated by demonstrating and analyzing the possible existence of oscillatory behaviors when considering an extended open configuration of the enzymatic network. Notably, we construct bifurcation diagrams that specify key parameters values controlling the co-existence of stable steady and non-steady oscillatory proteolytic dynamics.

Background Cellular response to changes in the concentration of different chemical species in the extracellular medium is induced by ligand binding to dedicated transmembrane receptors. Receptor density, distribution, and clustering may be key spatial features that influence effective and proper physical and biochemical cellular responses to many regulatory signals. Classical equations describing this kind of binding kinetics assume the distributions of interacting species to be homogeneous, neglecting by doing so the impact of clustering. As there is experimental evidence that receptors tend to group in clusters inside membrane domains, we investigated the effects of receptor clustering on cellular receptor ligand binding. Results We implemented a model of receptor binding using a Monte-Carlo algorithm to simulate ligand diffusion and binding. In some simple cases, analytic solutions for binding equilibrium of ligand on clusters of receptors are provided, and supported by simulation results. Our simulations show that the so-called "apparent" affinity of the ligand for the receptor decreases with clustering although the microscopic affinity remains constant. Conclusions Changing membrane receptors clustering could be a simple mechanism that allows cells to change and adapt its affinity/sensitivity toward a given stimulus.

Motivation: When comparing the organization of two genomes, it is important not to draw conclusions on their modes of evolution from a single most parsimonious scenario explaining their differences. Better estimations can be obtained by sampling many different genomic rearrangement scenarios. For this problem, the Double Cut and Join (DCJ) model, while less relevant, is computationally easier than the Hannenhalli–Pevzner (HP) model. Indeed, in some special cases, the total number of DCJ sorting scenarios can be analytically calculated, and uniformly distributed random DCJ scenarios can be drawn in polynomial running time, while the complexity of counting the number of HP scenarios and sampling from the uniform distribution of their space is unknown, and conjectured to be #P-complete. Statistical methods, like Markov chain Monte Carlo (MCMC) for sampling from the uniform distribution of the most parsimonious or the Bayesian distribution of all possible HP scenarios are required. Results: We use the computational facilities of the DCJ model to draw a sampling of HP scenarios. It is based on a parallel MCMC method that cools down DCJ scenarios to HP scenarios. We introduce two theorems underlying the theoretical mixing properties of this parallel MCMC method. The method was tested on yeast and mammalian genomic data, and allowed us to provide estimates of the different modes of evolution in diverse lineages. Availability: The program implemented in Java 1.5 programming language is available from http://www.renyi.hu/~miklosi/DCJ2HP/. Contact: miklosi@renyi.hu

The acid/base properties of proteins are essential in biochemistry, and proton binding is a valuable reporter on electrostatic interactions. We propose a constant-pH Monte Carlo strategy to compute protonation free energies and pKas. The solvent is described implicitly, through a generalized Born model. The electronic polarizability and backbone motions of the protein are included through the protein dielectric constant. Side chain motions are described explicitly, by the Monte Carlo scheme. An efficient computational algorithm is described, which allows us to treat the fluctuating shape of the protein/solvent boundary in a way that is numerically exact (within the GB framework) this contrasts with several previous constant-pH approaches. For a test set of six proteins and 78 titratable groups, the model performs well, with an rms error of 1.2 pH units. While this is slightly greater than a simple Null model (rms error of 1.1) and a fully empirical model (rms error of 0.9), it is obtained using physically meaningful model parameters, including a low protein dielectric of four. Importantly, similar performance is obtained for side chains with large and small pKa shifts (relative to a standard model compound). The titration curve slopes and the conformations sampled are reasonable. Several directions to improve the method further are discussed. © 2010 American Chemical Society. 114 32 10634 10648 Cited By :20

International audience; In (18)F-FDG PET, tumors are often characterized by their metabolically active volume and standardized uptake value (SUV). However, many approaches have been proposed to estimate tumor volume and SUV from (18)F-FDG PET images, none of them being widely agreed upon. We assessed the accuracy and robustness of 5 methods for tumor volume estimates and of 10 methods for SUV estimates in a large variety of configurations. METHODS: PET acquisitions of an anthropomorphic phantom containing 17 spheres (volumes between 0.43 and 97 mL, sphere-to-surrounding-activity concentration ratios between 2 and 68) were used. Forty-one nonspheric tumors (volumes between 0.6 and 92 mL, SUV of 2, 4, and 8) were also simulated and inserted in a real patient (18)F-FDG PET scan. Four threshold-based methods (including one, T(bgd), accounting for background activity) and a model-based method (Fit) described in the literature were used for tumor volume measurements. The mean SUV in the resulting volumes were calculated, without and with partial-volume effect (PVE) correction, as well as the maximum SUV (SUV(max)). The parameters involved in the tumor segmentation and SUV estimation methods were optimized using 3 approaches, corresponding to getting the best of each method or testing each method in more realistic situations in which the parameters cannot be perfectly optimized. RESULTS: In the phantom and simulated data, the T(bgd) and Fit methods yielded the most accurate volume estimates, with mean errors of 2% +/- 11% and -8% +/- 21% in the most realistic situations. Considering the simulated data, all SUV not corrected for PVE had a mean bias between -31% and -46%, much larger than the bias observed with SUV(max) (-11% +/- 23%) or with the PVE-corrected SUV based on T(bgd) and Fit (-2% +/- 10% and 3% +/- 24%). CONCLUSION: The method used to estimate tumor volume and SUV greatly affects the reliability of the estimates. The T(bgd) and Fit methods yielded low errors in volume estimates in a broad range of situations. The PVE-corrected SUV based on T(bgd) and Fit were more accurate and reproducible than SUV(max).

International audience; PURPOSE: Synchrotron microbeam radiation therapy (MRT) relies on spatial fractionation of the incident photon beam into parallel micron-wide beams. Our aim was to analyze the effects of MRT on normal brain and 9L gliosarcoma tissues, particularly on blood vessels. METHODS AND MATERIALS: Responses to MRT (two arrays, one lateral, one anteroposterior (2 × 400 Gy), intersecting orthogonally in the tumor region) were studied during 6 weeks using MRI, immunohistochemistry, and vascular endothelial growth factor Western blot. RESULTS: MRT increased the median survival time of irradiated rats (×3.25), significantly increased blood vessel permeability, and inhibited tumor growth; a cytotoxic effect on 9L cells was detected 5 days after irradiation. Significant decreases in tumoral blood volume fraction and vessel diameter were measured from 8 days after irradiation, due to loss of endothelial cells in tumors as detected by immunochemistry. Edema was observed in the normal brain exposed to both crossfired arrays about 6 weeks after irradiation. This edema was associated with changes in blood vessel morphology and an overexpression of vascular endothelial growth factor. Conversely, vascular parameters and vessel morphology in brain regions exposed to one of the two arrays were not damaged, and there was no loss of vascular endothelia. CONCLUSIONS: We show for the first time that preferential damage of MRT to tumor vessels versus preservation of radioresistant normal brain vessels contributes to the efficient palliation of 9L gliosarcomas in rats. Molecular pathways of repair mechanisms in normal and tumoral vascular networks after MRT may be essential for the improvement of such differential effects on the vasculature.

Background Biological networks are highly dynamic in response to environmental and physiological cues. This variability is in contrast to conventional analyses of biological networks, which have overwhelmingly employed static graph models which stay constant over time to describe biological systems and their underlying molecular interactions. Methods To overcome these limitations, we propose here a new statistical modelling framework, the ARTIVA formalism (Auto Regressive TIme VArying models), and an associated inferential procedure that allows us to learn temporally varying gene-regulation networks from biological time-course expression data. ARTIVA simultaneously infers the topology of a regulatory network and how it changes over time. It allows us to recover the chronology of regulatory associations for individual genes involved in a specific biological process (development, stress response, etc.). Results We demonstrate that the ARTIVA approach generates detailed insights into the function and dynamics of complex biological systems and exploits efficiently time-course data in systems biology. In particular, two biological scenarios are analyzed: the developmental stages of Drosophila melanogaster and the response of Saccharomyces cerevisiae to benomyl poisoning. Conclusions ARTIVA does recover essential temporal dependencies in biological systems from transcriptional data, and provide a natural starting point to learn and investigate their dynamics in greater detail.

The extent to which self-adopted or intervention-related changes in behaviors affect the course of epidemics remains a key issue for outbreak control. This study attempted to quantify the effect of such changes on the risk of infection in different settings, i.e., the community and hospitals. The 2002-2003 severe acute respiratory syndrome (SARS) outbreak in Hong Kong, where 27% of cases were healthcare workers, was used as an example. A stochastic compartmental SEIR (susceptible-exposed-infectious-removed) model was used: the population was split into healthcare workers, hospitalized people and general population. Super spreading events (SSEs) were taken into account in the model. The temporal evolutions of the daily effective contact rates in the community and hospitals were modeled with smooth functions. Data augmentation techniques and Markov chain Monte Carlo (MCMC) methods were applied to estimate SARS epidemiological parameters. In particular, estimates of daily reproduction numbers were provided for each subpopulation. The average duration of the SARS infectious period was estimated to be 9.3 days (±0.3 days). The model was able to disentangle the impact of the two SSEs from background transmission rates. The effective contact rates, which were estimated on a daily basis, decreased with time, reaching zero inside hospitals. This observation suggests that public health measures and possible changes in individual behaviors effectively reduced transmission, especially in hospitals. The temporal patterns of reproduction numbers were similar for healthcare workers and the general population, indicating that on average, an infectious healthcare worker did not infect more people than any other infectious person. We provide a general method to estimate time dependence of parameters in structured epidemic models, which enables investigation of the impact of control measures and behavioral changes in different settings. © 2009 Cori et al., published_or_final_version

International audience; Analysis of count data from clinical trials using mixed effect analysis has recently become widely used. However, algorithms available for the parameter estimation, including LAPLACE and Gaussian quadrature (GQ), are associated with certain limitations, including bias in parameter estimates and the long analysis runtime. The stochastic approximation expectation maximization (SAEM) algorithm has proven to be a very efficient and powerful tool in the analysis of continuous data. The aim of this study was to implement and investigate the performance of a new SAEM algorithm for application to count data. A new SAEM algorithm was implemented in MATLAB for estimation of both, parameters and the Fisher information matrix. Stochastic Monte Carlo simulations followed by re-estimation were performed according to scenarios used in previous studies (part I) to investigate properties of alternative algorithms (Plan et al., 2008, Abstr 1372 [ http://wwwpage-meetingorg/?abstract=1372 ]). A single scenario was used to explore six probability distribution models. For parameter estimation, the relative bias was less than 0.92% and 4.13% for fixed and random effects, for all models studied including ones accounting for over- or under-dispersion. Empirical and estimated relative standard errors were similar, with distance between them being

International audience; This article focuses on parameter estimation of multilevel nonlinear mixed-effects models (MNLMEMs). These models are used to analyze data presenting multiple hierarchical levels of grouping (cluster data, clinical trials with several observation periods, ...). The variability of the individual parameters of the regression function is thus decomposed as a between-subject variability and higher levels of variability (e.g. within-subject variability). We propose maximum likelihood estimates of parameters of those MNLMEMs with 2 levels of random effects, using an extension of the stochastic approximation version of expectation-maximization (SAEM)-Monte Carlo Markov chain algorithm. The extended SAEM algorithm is split into an explicit direct expectation-maximization (EM) algorithm and a stochastic EM part. Compared to the original algorithm, additional sufficient statistics have to be approximated by relying on the conditional distribution of the second level of random effects. This estimation method is evaluated on pharmacokinetic crossover simulated trials, mimicking theophylline concentration data. Results obtained on those data sets with either the SAEM algorithm or the first-order conditional estimates (FOCE) algorithm (implemented in the nlme function of R software) are compared: biases and root mean square errors of almost all the SAEM estimates are smaller than the FOCE ones. Finally, we apply the extended SAEM algorithm to analyze the pharmacokinetic interaction of tenofovir on atazanavir, a novel protease inhibitor, from the Agence Nationale de Recherche sur le Sida 107-Puzzle 2 study. A significant decrease of the area under the curve of atazanavir is found in patients receiving both treatments.

In nonlinear mixed-effects models, estimation methods based on a linearization of the likelihood are widely used although they have several methodological drawbacks. Kuhn and Lavielle (Comput. Statist. Data Anal. 49:1020-1038 (2005)) developed an estimation method which combines the SAEM (Stochastic Approximation EM) algorithm, with a MCMC (Markov Chain Monte Carlo) procedure for maximum likelihood estimation in nonlinear mixed-effects models without linearization. This method is implemented in the Matlab software MONOLIX which is available at http://www.math.u-psud.fr/~lavielle/monolix/logiciels. In this paper we apply MONOLIX to the analysis of the pharmacokinetics of saquinavir, a protease inhibitor, from concentrations measured after single dose administration in 100 HIV patients, some with advance disease. We also illustrate how to use MONOLIX to build the covariate model using the Bayesian Information Criterion. Saquinavir oral clearance (CL/F) was estimated to be 1.26 L/h and to increase with body mass index, the inter-patient variability for CL/F being 120%. Several methodological developments are ongoing to extend SAEM which is a very promising estimation method for population pharmacockinetic/pharmacodynamic analyses.

Model evaluation is an important issue in population analyses. We aimed to perform a systematic review of all population pharmacokinetic and/or pharmacodynamic analyses published between 2002 and 2004 to survey the current methods used to evaluate models and to assess whether those models were adequately evaluated. We selected 324 articles in MEDLINE using defined key words and built a data abstraction form composed of a checklist of items to extract the relevant information from these articles with respect to model evaluation. In the data abstraction form, evaluation methods were divided into three subsections: basic internal methods (goodness-of-fit [GOF] plots, uncertainty in parameter estimates and model sensitivity), advanced internal methods (data splitting, resampling techniques and Monte Carlo simulations) and external model evaluation. Basic internal evaluation was the most frequently described method in the reports: 65% of the models involved GOF evaluation. Standard errors or confidence intervals were reported for 50% of fixed effects but only for 22% of random effects. Advanced internal methods were used in approximately 25% of models: data splitting was more often used than bootstrap and cross-validation; simulations were used in 6% of models to evaluate models by a visual predictive check or by a posterior predictive check. External evaluation was performed in only 7% of models. Using the subjective synthesis of model evaluation for each article, we judged the models to be adequately evaluated in 28% of pharmacokinetic models and 26% of pharmacodynamic models. Basic internal evaluation was preferred to more advanced methods, probably because the former is performed easily with most software. We also noticed that when the aim of modelling was predictive, advanced internal methods or more stringent methods were more often used.

Model evaluation is an important issue in population analyses. We aimed to perform a systematic review of all population pharmacokinetic and/or pharmacodynamic analyses published between 2002 and 2004 to survey the current methods used to evaluate models and to assess whether those models were adequately evaluated. We selected 324 articles in MEDLINE using defined key words and built a data abstraction form composed of a checklist of items to extract the relevant information from these articles with respect to model evaluation. In the data abstraction form, evaluation methods were divided into three subsections: basic internal methods (goodness-of-fit [GOF] plots, uncertainty in parameter estimates and model sensitivity), advanced internal methods (data splitting, resampling techniques and Monte Carlo simulations) and external model evaluation. Basic internal evaluation was the most frequently described method in the reports: 65% of the models involved GOF evaluation. Standard errors or confidence intervals were reported for 50% of fixed effects but only for 22% of random effects. Advanced internal methods were used in approximately 25% of models: data splitting was more often used than bootstrap and cross-validation; simulations were used in 6% of models to evaluate models by a visual predictive check or by a posterior predictive check. External evaluation was performed in only 7% of models. Using the subjective synthesis of model evaluation for each article, we judged the models to be adequately evaluated in 28% of pharmacokinetic models and 26% of pharmacodynamic models. Basic internal evaluation was preferred to more advanced methods, probably because the former is performed easily with most software. We also noticed that when the aim of modelling was predictive, advanced internal methods or more stringent methods were more often used.

International audience; Exposure to lead is a special problem in children, because they are more highly exposed than adults and because this pollutant, which accumulates in the body, induces neurobehavioral and cognitive effects. The objective of this study was to determine the probability density of the lead exposure dose of a 2-year-old child around an old mine site and to analyze its uncertainties, especially those associated with the bioavailability of lead in soil. Children's exposure was estimated indirectly from environmental samples (soils, domestic dust, water, air) and parameters (volume inhaled, body weight, soil intake rate, water intake, dietary intake) from the literature. Uncertainty and variability were analyzed separately in a two-dimensional Monte Carlo simulation with Crystal Ball software. Exposure doses were simulated with different methods for accessing the bioavailability of lead in soil. The exposure dose per kilogram of body weight varied from 2 microg/kgday at the 5th percentile to 5.5 microg/kgday at the 95th percentile (and from 2 to 10 microg/kgday, respectively, when ignoring bioavailability). The principal factors of variation were dietary intake, soil concentrations, and soil ingestion. The principal uncertainties were associated with the level of soil ingestion and the bioavailability of lead. Reducing uncertainty about the bioavailability of lead in soil by taking into account information about the type of mineral made it possible to increase our degree of confidence (from 25% to more than 95%) that the median exposure dose does not exceed the Tolerable Daily Intake. Knowledge of the mineral very substantially increases the degree of confidence in estimates of children's lead exposure around an old mining site by reducing the uncertainty associated with lead's bioavailability.

Reliable estimation methods for non-linear mixed-effects models are now available and, although these models are increasingly used, only a limited number of statistical developments for their evaluation have been reported. We develop a criterion and a test to evaluate nonlinear mixed-effects models based on the whole predictive distribution. For each observation, we define the prediction discrepancy (pd) as the percentile of the observation in the whole marginal predictive distribution under H(0). We propose to compute prediction discrepancies using Monte Carlo integration which does not require model approximation. If the model is valid, these pd should be uniformly distributed over (0, 1) which can be tested by a Kolmogorov-Smirnov test. In a simulation study based on a standard population pharmacokinetic model, we compare and show the interest of this criterion with respect to the one most frequently used to evaluate nonlinear mixed-effects models: standardized prediction errors (spe) which are evaluated using a first order approximation of the model. Trends in pd can also be evaluated via several plots to check for specific departures from the model.

Most studies of neighborhood effects on health have used the multilevel approach. However, since this methodology does not incorporate any notion of space, it may not provide optimal epidemiologic information when modeling variations or when investigating associations between contextual factors and health. Investigating mental disorders due to psychoactive substance use among all 65,830 individuals aged 40-59 years in 2001 in Malmö, Sweden, geolocated at their place of residence, the authors compared a spatial analytical perspective, which builds notions of space into hypotheses and methods, with the multilevel approach. Geoadditive models provided precise cartographic information on spatial variations in prevalence independent of administrative boundaries. The multilevel model showed significant neighborhood variations in the prevalence of substance-related disorders. However, hierarchical geostatistical models provided information on not only the magnitude but also the scale of neighborhood variations, indicating a significant correlation between neighborhoods in close proximity to each other. The prevalence of disorders increased with neighborhood deprivation. Far stronger associations were observed when using indicators measured in spatially adaptive areas, centered on residences of individuals, smaller in size than administrative neighborhoods. In neighborhood studies, building notions of space into analytical procedures may yield more comprehensive information than heretofore has been gathered on the spatial distribution of outcomes.

Recently, regression analysis of the cumulative incidence function has gained interest in competing risks data analysis, through the model proposed by Fine and Gray (JASA 1999; 94: 496-509). In this note, we point out that inclusion of time-dependent covariates in this model can lead to serious bias. We illustrate the problems arising in such a context, using bone marrow transplant data as a working example and numerical simulations. Practical advices are given, preventing the misuse of this model.

Background In order to detect potential disease clusters where a putative source cannot be specified, classical procedures scan the geographical area with circular windows through a specified grid imposed to the map. However, the choice of the windows' shapes, sizes and centers is critical and different choices may not provide exactly the same results. The aim of our work was to use an Oblique Decision Tree model (ODT) which provides potential clusters without pre-specifying shapes, sizes or centers. For this purpose, we have developed an ODT-algorithm to find an oblique partition of the space defined by the geographic coordinates. Methods ODT is based on the classification and regression tree (CART). As CART finds out rectangular partitions of the covariate space, ODT provides oblique partitions maximizing the interclass variance of the independent variable. Since it is a NP-Hard problem in RN, classical ODT-algorithms use evolutionary procedures or heuristics. We have developed an optimal ODT-algorithm in R2, based on the directions defined by each couple of point locations. This partition provided potential clusters which can be tested with Monte-Carlo inference. We applied the ODT-model to a dataset in order to identify potential high risk clusters of malaria in a village in Western Africa during the dry season. The ODT results were compared with those of the Kulldorff' s SaTScan™. Results The ODT procedure provided four classes of risk of infection. In the first high risk class 60%, 95% confidence interval (CI95%) [52.22–67.55], of the children was infected. Monte-Carlo inference showed that the spatial pattern issued from the ODT-model was significant (p < 0.0001). Satscan results yielded one significant cluster where the risk of disease was high with an infectious rate of 54.21%, CI95% [47.51–60.75]. Obviously, his center was located within the first high risk ODT class. Both procedures provided similar results identifying a high risk cluster in the western part of the village where a mosquito breeding point was located. Conclusion ODT-models improve the classical scanning procedures by detecting potential disease clusters independently of any specification of the shapes, sizes or centers of the clusters.

BACKGROUND: Maternal age is an important factor in reproduction. Can assisted reproduction technologies (ART) fully compensate for the decline in fertility with age? METHODS: We used a computer simulation (Monte Carlo) model of reproduction, combining the monthly probabilities of conceiving, the risk of miscarriage and the probability of becoming age-dependently permanently sterile. RESULTS: Under natural conditions, 75% of women starting to try to conceive at age 30 years will have a conception ending in a live birth within 1 year, 66% at age 35 years and 44% at age 40 years. Within 4 years the success rates will be respectively 91, 84 and 64%. If women turn to ART after 4, 3 or 2 years respectively without conception, and if the rate of success is as observed after two cycles of insemination in IVF, ART makes up for only half of the births lost by postponing a first attempt of pregnancy from age 30 to 35 years, and

International audience; The unified neutral model of biodiversity proposed by S. Hubbell is solved analytically: The distributions of species abundance in the metacommunity and in a local community are calculated exactly as a function of speciation and migration rates and of the size of the community. In the limit of large population sizes the densities of species of given relative abundance are found to be given by universal functions depending only on two parameters.

International audience; Globular proteins have highly compact structures and the corresponding packing interactions are widely considered as the principal determinant of the native structure. It is therefore important that theoretical approaches to protein design explicitly take in account packing, which requires that a full atomic representation of the designed protein is maintained. As a first step towards this goal, we have developed in this report an inverse folding algorithm with the aim of specifically designing amino acid sequences which optimise sidechain packing for a given protein fold. The design is performed by a global Monte Carlo optimisation in sequence space, with constant amino acid composition and a full-atom representation of the various protein models. Packing is defined by a Lennard-Jones potential. The program was tested by designing stable sequence variants for the chymotrypsin inhibitor fold. The final protein models showed an increase in intramolecular atomic contacts and a decrease in the overall volume compared to the native structure. Starting from the backbone only of the target structure, the algorithm did gradually retrieve reliable though limited sequence information. Higher compatibility might be achieved by improving the potential, however our results suggest that packing interactions are an essential element of a yet-to-be-defined successful energy function for protein design.