Data-driven clustering : new methods and applications

Abstract : This work is concerned with the development and application of novel unsupervised learning methods, having in mind two target applications: the analysis of forensic case data and the classification of remote sensing images. First, a method based on a symbolic optimization of the inter-sample distance measure is proposed to improve the flexibility of spectral clustering algorithms, and applied to the problem of forensic case data. This distance is optimized using a loss function related to the preservation of neighborhood structure between the input space and the space of principal components, and solutions are found using genetic programming. Results are compared to a variety of state-of--the-art clustering algorithms. Subsequently, a new large-scale clustering method based on a joint optimization of feature extraction and classification is proposed and applied to various databases, including two hyperspectral remote sensing images. The algorithm makes uses of a functional model (e.g., a neural network) for clustering which is trained by stochastic gradient descent. Results indicate that such a technique can easily scale to huge databases, can avoid the so-called out-of-sample problem, and can compete with or even outperform existing clustering algorithms on both artificial data and real remote sensing images. This is verified on small databases as well as very large problems. Résumé : Ce travail de recherche porte sur le développement et l'application de méthodes d'apprentissage dites non supervisées. Les applications visées par ces méthodes sont l'analyse de données forensiques et la classification d'images hyperspectrales en télédétection. Dans un premier temps, une méthodologie de classification non supervisée fondée sur l'optimisation symbolique d'une mesure de distance inter-échantillons est proposée. Cette mesure est obtenue en optimisant une fonction de coût reliée à la préservation de la structure de voisinage d'un point entre l'espace des variables initiales et l'espace des composantes principales. Cette méthode est appliquée à l'analyse de données forensiques et comparée à un éventail de méthodes déjà existantes. En second lieu, une méthode fondée sur une optimisation conjointe des tâches de sélection de variables et de classification est implémentée dans un réseau de neurones et appliquée à diverses bases de données, dont deux images hyperspectrales. Le réseau de neurones est entraîné à l'aide d'un algorithme de gradient stochastique, ce qui rend cette technique applicable à des images de très haute résolution. Les résultats de l'application de cette dernière montrent que l'utilisation d'une telle technique permet de classifier de très grandes bases de données sans difficulté et donne des résultats avantageusement comparables aux méthodes existantes.

Comparative performance of supertree algorithms in large data sets using the soapberry family (Sapindaceae) as a case study.

For the last 2 decades, supertree reconstruction has been an active field of research and has seen the development of a large number of major algorithms. Because of the growing popularity of the supertree methods, it has become necessary to evaluate the performance of these algorithms to determine which are the best options (especially with regard to the supermatrix approach that is widely used). In this study, seven of the most commonly used supertree methods are investigated by using a large empirical data set (in terms of number of taxa and molecular markers) from the worldwide flowering plant family Sapindaceae. Supertree methods were evaluated using several criteria: similarity of the supertrees with the input trees, similarity between the supertrees and the total evidence tree, level of resolution of the supertree and computational time required by the algorithm. Additional analyses were also conducted on a reduced data set to test if the performance levels were affected by the heuristic searches rather than the algorithms themselves. Based on our results, two main groups of supertree methods were identified: on one hand, the matrix representation with parsimony (MRP), MinFlip, and MinCut methods performed well according to our criteria, whereas the average consensus, split fit, and most similar supertree methods showed a poorer performance or at least did not behave the same way as the total evidence tree. Results for the super distance matrix, that is, the most recent approach tested here, were promising with at least one derived method performing as well as MRP, MinFlip, and MinCut. The output of each method was only slightly improved when applied to the reduced data set, suggesting a correct behavior of the heuristic searches and a relatively low sensitivity of the algorithms to data set sizes and missing data. Results also showed that the MRP analyses could reach a high level of quality even when using a simple heuristic search strategy, with the exception of MRP with Purvis coding scheme and reversible parsimony. The future of supertrees lies in the implementation of a standardized heuristic search for all methods and the increase in computing power to handle large data sets. The latter would prove to be particularly useful for promising approaches such as the maximum quartet fit method that yet requires substantial computing power.

Prevalence of Neural Tube Defects (NTD) in Younger Mothers in Europe 2000-2008: Analysis of the EUROCAT Database - Report to Bayer Schiering Pharma

M.C. Addor is included in the Eurocat Working Group

Neural networks as mechanisms to regulate division of labor.

Abstract In social insects, workers perform a multitude of tasks, such as foraging, nest construction, and brood rearing, without central control of how work is allocated among individuals. It has been suggested that workers choose a task by responding to stimuli gathered from the environment. Response-threshold models assume that individuals in a colony vary in the stimulus intensity (response threshold) at which they begin to perform the corresponding task. Here we highlight the limitations of these models with respect to colony performance in task allocation. First, we show with analysis and quantitative simulations that the deterministic response-threshold model constrains the workers' behavioral flexibility under some stimulus conditions. Next, we show that the probabilistic response-threshold model fails to explain precise colony responses to varying stimuli. Both of these limitations would be detrimental to colony performance when dynamic and precise task allocation is needed. To address these problems, we propose extensions of the response-threshold model by adding variables that weigh stimuli. We test the extended response-threshold model in a foraging scenario and show in simulations that it results in an efficient task allocation. Finally, we show that response-threshold models can be formulated as artificial neural networks, which consequently provide a comprehensive framework for modeling task allocation in social insects.

Neural differentiation and therapeutic potential of adipose tissue derived stem cells.

Neural tissue has historically been regarded as having poor regenerative capacity but recent advances in the growing fields of tissue engineering and regenerative medicine have opened new hopes for the treatment of nerve injuries and neurodegenerative disorders. Adipose tissue has been shown to contain a large quantity of adult stem cells (ASC). These cells can be easily harvested with low associated morbidity and because of their potential to differentiate into multiple cell types, their use has been suggested for a wide variety of therapeutic applications. In this review we examine the evidence indicating that ASC can stimulate nerve regeneration by both undergoing neural differentiation and through the release of a range of growth factors. We also discuss some of the issues that need to be addressed before ASC can be developed as an effective cellular therapy for the treatment of neural tissue disorders.

Machine learning analysis and modeling of interest rate curves

The present research deals with the review of the analysis and modeling of Swiss franc interest rate curves (IRC) by using unsupervised (SOM, Gaussian Mixtures) and supervised machine (MLP) learning algorithms. IRC are considered as objects embedded into different feature spaces: maturities; maturity-date, parameters of Nelson-Siegel model (NSM). Analysis of NSM parameters and their temporal and clustering structures helps to understand the relevance of model and its potential use for the forecasting. Mapping of IRC in a maturity-date feature space is presented and analyzed for the visualization and forecasting purposes.

Advanced geostatistical and machine-learning models for spatial data analysis of radioactively contaminated regions

Radioactive soil-contamination mapping and risk assessment is a vital issue for decision makers. Traditional approaches for mapping the spatial concentration of radionuclides employ various regression-based models, which usually provide a single-value prediction realization accompanied (in some cases) by estimation error. Such approaches do not provide the capability for rigorous uncertainty quantification or probabilistic mapping. Machine learning is a recent and fast-developing approach based on learning patterns and information from data. Artificial neural networks for prediction mapping have been especially powerful in combination with spatial statistics. A data-driven approach provides the opportunity to integrate additional relevant information about spatial phenomena into a prediction model for more accurate spatial estimates and associated uncertainty. Machine-learning algorithms can also be used for a wider spectrum of problems than before: classification, probability density estimation, and so forth. Stochastic simulations are used to model spatial variability and uncertainty. Unlike regression models, they provide multiple realizations of a particular spatial pattern that allow uncertainty and risk quantification. This paper reviews the most recent methods of spatial data analysis, prediction, and risk mapping, based on machine learning and stochastic simulations in comparison with more traditional regression models. The radioactive fallout from the Chernobyl Nuclear Power Plant accident is used to illustrate the application of the models for prediction and classification problems. This fallout is a unique case study that provides the challenging task of analyzing huge amounts of data ('hard' direct measurements, as well as supplementary information and expert estimates) and solving particular decision-oriented problems.

Etude de la variation interlaboratoire du test de dépistage prénatal des défauts de fermeture du tube neural [Interlaboratory variation in the prenatal screening test for neural tube closing defects].

Early detection of neural-tude defects is possible by determining Alpha-fetoprotein (AFP) in maternal serum. 16'685 pregnant women were observed. Three methods for the determination of the "normal" range are compared. The first one, already used in similar studies, makes use of a constant multiple of the median. The other two ones make use of robust estimates of location and scale. Their comparison shows the interest of the robust methods to reduce the interlaboratory variability.

A circuit-based gatekeeper for adult neural stem cell proliferation: Parvalbumin-expressing interneurons of the dentate gyrus control the activation and proliferation of quiescent adult neural stem cells.

Newborn neurons are generated in the adult hippocampus from a pool of self-renewing stem cells located in the subgranular zone (SGZ) of the dentate gyrus. Their activation, proliferation, and maturation depend on a host of environmental and cellular factors but, until recently, the contribution of local neuronal circuitry to this process was relatively unknown. In their recent publication, Song and colleagues have uncovered a novel circuit-based mechanism by which release of the neurotransmitter, γ-aminobutyric acid (GABA), from parvalbumin-expressing (PV) interneurons, can hold radial glia-like (RGL) stem cells of the adult SGZ in a quiescent state. This tonic GABAergic signal, dependent upon the activation of γ(2) subunit-containing GABA(A) receptors of RGL stem cells, can thus prevent their proliferation and subsequent maturation or return them to quiescence if previously activated. PV interneurons are thus capable of suppressing neurogenesis during periods of high network activity and facilitating neurogenesis when network activity is low.

Coagulation Factor Xa Activates Thrombin and Signals Ischemic Neuronal Death Via Par-1 and JNK in Ischemic Neural Tissue

Background: The coagulation factor thrombin mediates ischemic neuronal deathand, at a low concentration, induces tolerance to ischemia.We investigated its modeof activation in ischemic neural tissue using an in vitro approach to distinguish therole of circulating coagulation factors from endogenous cerebral mechanisms. Wealso studied the signalling pathway downstream of thrombin in ischemia and afterthrombin preconditioning.Methods: Rat organotypic hippocampal slice cultures to 30 minute oxygen (5%)and glucose (1 mmol/L) deprivation (OGD).Results: Selective factor Xa (FXa) inhibition by fondaparinux during and afterOGD significantly reduced neuronal death in the CA1 after 48 hours. Thrombinactivity was increased in the medium 24 hours after OGD and this increasewas prevented by fondaparinux suggesting that FXa catalyzes the conversion ofprothrombin to thrombin in neural tissue after ischemia in vitro. Treatment withSCH79797, a selective antagonist of the thrombin receptor protease activatedreceptor-1 (PAR-1), significantly decreased neuronal cell death indicating thatthrombin signals ischemic damage via PAR-1. The JNK pathway plays an importantrole in cerebral ischemia and we observed activation of the JNK substrate,c-Jun in our model. Both the FXa inhibitor, fondaparinux and the PAR-1 antagonistSCH79797, decreased the level of phospho-c-Jun Ser73. After thrombin preconditioningc-Jun was activated by phosphorylation in the nuclei of neurons of the CA1.Treatment with a synthetic thrombin receptor agonist resulted in the same c-Junactivation profile and protection against subsequent OGD indicating that thrombinalso signals via PAR-1 and c-Jun in cell protection.Conclusion: These results indicate that FXa activates thrombin in cerebral ischemia,leading via PAR-1 to the activation of the JNK pathway resulting in neuronal death.Thrombin induced tolerance also involves PAR-1 and JNK, revealing commonfeatures in cell death and survival signalling.

Stochastic models and numerical algorithms for a class of regulatory gene networks.

Regulatory gene networks contain generic modules, like those involving feedback loops, which are essential for the regulation of many biological functions (Guido et al. in Nature 439:856-860, 2006). We consider a class of self-regulated genes which are the building blocks of many regulatory gene networks, and study the steady-state distribution of the associated Gillespie algorithm by providing efficient numerical algorithms. We also study a regulatory gene network of interest in gene therapy, using mean-field models with time delays. Convergence of the related time-nonhomogeneous Markov chain is established for a class of linear catalytic networks with feedback loops.

The long-term survival of in vitro engineered nervous tissue derived from the specific neural differentiation of mouse embryonic stem cells.

Embryonic stem cells (ESCs) offer attractive prospective as potential source of neurons for cell replacement therapy in human neurodegenerative diseases. Besides, ESCs neural differentiation enables in vitro tissue engineering for fundamental research and drug discovery aimed at the nervous system. We have established stable and long-term three-dimensional (3D) culture conditions which can be used to model long latency and complex neurodegenerative diseases. Mouse ESCs-derived neural progenitor cells generated by MS5 stromal cells induction, result in strictly neural 3D cultures of about 120-mum thick, whose cells expressed mature neuronal, astrocytes and myelin markers. Neurons were from the glutamatergic and gabaergic lineages. This nervous tissue was spatially organized in specific layers resembling brain sub-ependymal (SE) nervous tissue, and was maintained in vitro for at least 3.5 months with great stability. Electron microscopy showed the presence of mature synapses and myelinated axons, suggesting functional maturation. Electrophysiological activity revealed biological signals involving action potential propagation along neuronal fibres and synaptic-like release of neurotransmitters. The rapid development and stabilization of this 3D cultures model result in an abundant and long-lasting production that is compatible with multiple and productive investigations for neurodegenerative diseases modeling, drug and toxicology screening, stress and aging research.

Stimulus statistics shape oscillations in nonlinear recurrent neural networks.

Rhythmic activity plays a central role in neural computations and brain functions ranging from homeostasis to attention, as well as in neurological and neuropsychiatric disorders. Despite this pervasiveness, little is known about the mechanisms whereby the frequency and power of oscillatory activity are modulated, and how they reflect the inputs received by neurons. Numerous studies have reported input-dependent fluctuations in peak frequency and power (as well as couplings across these features). However, it remains unresolved what mediates these spectral shifts among neural populations. Extending previous findings regarding stochastic nonlinear systems and experimental observations, we provide analytical insights regarding oscillatory responses of neural populations to stimulation from either endogenous or exogenous origins. Using a deceptively simple yet sparse and randomly connected network of neurons, we show how spiking inputs can reliably modulate the peak frequency and power expressed by synchronous neural populations without any changes in circuitry. Our results reveal that a generic, non-nonlinear and input-induced mechanism can robustly mediate these spectral fluctuations, and thus provide a framework in which inputs to the neurons bidirectionally regulate both the frequency and power expressed by synchronous populations. Theoretical and computational analysis of the ensuing spectral fluctuations was found to reflect the underlying dynamics of the input stimuli driving the neurons. Our results provide insights regarding a generic mechanism supporting spectral transitions observed across cortical networks and spanning multiple frequency bands.

Citrobacter rodentium Requires Intestinal Neural Wiskott-Aldrich Syndrome Protein (N-WASp) for Actin Pedestal Formation, Colonization and Epithelial Barrier Disruption In Vivo

Background: Citrobacter rodentium is a natural mouse pathogen that is genetically closelyrelated to the human enteric pathogens enteropathogenic and enterohemorrhagic E. coli.Among the repertoire of conserved virulence factors that these pathogens deliver via typeIII secretion, Tir and EspF are responsible for the formation of characteristic actin-richpedestals and disruption of tight junction integrity, respectively. There is evidence In Vitrothese effectors accomplish this, at least in part, by subverting the normal host cellularfunctions of N-WASP, a critical regulator of branched chain actin assembly. Although NWASPhas been shown to be involved in pedestal formation In Vitro, the requirements ofN-WASP-mediated actin pedestals for intestinal colonization by attaching/effacing (A/E)pathogens In Vivo is not known. Furthermore, it is not known whether N-WASP is requiredfor EspF-mediated tight junction disruption. Methods: To investigate the role of N-WASPin the gut epithelium, we generated mice with intestine-specific deletion of N-WASP(iNWKO), by mating mice homozygous for a floxed N-WASP allele (N-WASPL2L/L2L) tomice expressing Cre recombinase under the villin promoter. Separately housed groups ofWT and iNWKO mice were inoculated with 5x108 GFP-expressing C. rodentium by intragastriclavage. Stool was collected 2, 4, 7, and 12 days after infection, and recoverablecolony forming units (CFUs) of C. rodentium were quantified by plating serial dilutions ofhomogenized stool on MacConkey's agar. GFP+ colonies were counted after 24 hoursincubation at 37°C. The presence of actin pedestals was investigated by electron microscopy(EM), and tight junction morphology was assessed by immunofluorescence staining ofoccludin, ZO-1 and claudin-2. Results: C. rodentium infection did not result in mortalityin WT or iNWKO mice. Compared to controls, iNWKO mice exhibited higher levels ofbacterial shedding during the first 4 days of infection (day 4 average: WT 5.2x104 CFU/gvs. iNWKO 4.7x105 CFU/g, p=0.08), followed by a more rapid clearance of C. rodentium, (day7-12 average: WT 2x106 CFU/g vs. iNWKO 2.7x105, p=0.01). EM and immunofluorescencerevealed the complete lack of actin pedestals in iNWKO mice and no mucosa-associatedGFP+ C. rodentium by day 7. WT controls exhibited tight junction disruption, reflected byaltered distribution of ZO-1, whereas iNWKO mice had no change in the pattern of ZO-1.Conclusion: Intestinal N-WASP is required for actin pedestal formation by C. rodentium InVivo, and ablation of N-WASP is associated with more rapid bacterial clearance and decreasedability of C. rodentium to disrupt intercellular junctions.