To select or to weigh: a comparative study of linear combination schemes for superparent-one-dependence estimators

We conduct a large-scale comparative study on linearly combining superparent-one-dependence estimators (SPODEs), a popular family of seminaive Bayesian classifiers. Altogether, 16 model selection and weighing schemes, 58 benchmark data sets, and various statistical tests are employed. This paper's main contributions are threefold. First, it formally presents each scheme's definition, rationale, and time complexity and hence can serve as a comprehensive reference for researchers interested in ensemble learning. Second, it offers bias-variance analysis for each scheme's classification error performance. Third, it identifies effective schemes that meet various needs in practice. This leads to accurate and fast classification algorithms which have an immediate and significant impact on real-world applications. Another important feature of our study is using a variety of statistical tests to evaluate multiple learning methods across multiple data sets.

Analysis of linear networks with inconsistent initial conditions

This paper presents a new method to analyze timeinvariant linear networks allowing the existence of inconsistent initial conditions. This method is based on the use of distributions and state equations. Any time-invariant linear network can be analyzed. The network can involve any kind of pure or controlled sources. Also, the transferences of energy that occur at t=O are determined, and the concept of connection energy is introduced. The algorithms are easily implemented in a computer program.

Linear least squares estimation of the first order moving average parameter

We propose an iterative procedure to minimize the sum of squares function which avoids the nonlinear nature of estimating the first order moving average parameter and provides a closed form of the estimator. The asymptotic properties of the method are discussed and the consistency of the linear least squares estimator is proved for the invertible case. We perform various Monte Carlo experiments in order to compare the sample properties of the linear least squares estimator with its nonlinear counterpart for the conditional and unconditional cases. Some examples are also discussed

A theoretical and practical study on linear reforms of dual taxes

[cat] En aquest treball extenem les reformes lineals introduïdes per Pfähler (1984) al cas d’impostos duals. Estudiem l’efecte relatiu que els retalls lineals duals d’un impost dual tenen sobre la distribució de la desigualtat -es pot fer un estudi simètric per al cas d’augments d’impostos-. Tambe introduïm mesures del grau de progressivitat d’impostos duals i mostrem que estan connectades amb el criteri de dominació de Lorenz. Addicionalment, estudiem l’elasticitat de la càrrega fiscal de cadascuna de les reformes proposades. Finalment, gràcies a un model de microsimulació i una gran base de dades que conté informació sobre l’IRPF espanyol de l’any 2004, 1) comparem l’efecte que diferents reformes tindrien sobre l’impost dual espanyol i 2) estudiem quina redistribució de la riquesa va suposar la reforma dual de l’IRPF (Llei ’35/2006’) respecte l’anterior impost.

Drug-induced linear IgA bullous dermatosis probably induced by furosemide.

Linear IgA bullous dermatosis (LABD) is an autoimmune disease, characterized by linear deposition of IgA along the basement membrane zone. Drug-induced LABD is rare but increasing in frequency. A new case of drug-induced LABD associated with the administration of furosemide is described.

Enhanced pulse propagation in non-linear arrays of oscillators

The propagation of a pulse in a nonlinear array of oscillators is influenced by the nature of the array and by its coupling to a thermal environment. For example, in some arrays a pulse can be speeded up while in others a pulse can be slowed down by raising the temperature. We begin by showing that an energy pulse (one dimension) or energy front (two dimensions) travels more rapidly and remains more localized over greater distances in an isolated array (microcanonical) of hard springs than in a harmonic array or in a soft-springed array. Increasing the pulse amplitude causes it to speed up in a hard chain, leaves the pulse speed unchanged in a harmonic system, and slows down the pulse in a soft chain. Connection of each site to a thermal environment (canonical) affects these results very differently in each type of array. In a hard chain the dissipative forces slow down the pulse while raising the temperature speeds it up. In a soft chain the opposite occurs: the dissipative forces actually speed up the pulse, while raising the temperature slows it down. In a harmonic chain neither dissipation nor temperature changes affect the pulse speed. These and other results are explained on the basis of the frequency vs energy relations in the various arrays

Resting-State Functional Connectivity Emerges from Structurally and Dynamically Shaped Slow Linear Fluctuations.

Brain fluctuations at rest are not random but are structured in spatial patterns of correlated activity across different brain areas. The question of how resting-state functional connectivity (FC) emerges from the brain's anatomical connections has motivated several experimental and computational studies to understand structure-function relationships. However, the mechanistic origin of resting state is obscured by large-scale models' complexity, and a close structure-function relation is still an open problem. Thus, a realistic but simple enough description of relevant brain dynamics is needed. Here, we derived a dynamic mean field model that consistently summarizes the realistic dynamics of a detailed spiking and conductance-based synaptic large-scale network, in which connectivity is constrained by diffusion imaging data from human subjects. The dynamic mean field approximates the ensemble dynamics, whose temporal evolution is dominated by the longest time scale of the system. With this reduction, we demonstrated that FC emerges as structured linear fluctuations around a stable low firing activity state close to destabilization. Moreover, the model can be further and crucially simplified into a set of motion equations for statistical moments, providing a direct analytical link between anatomical structure, neural network dynamics, and FC. Our study suggests that FC arises from noise propagation and dynamical slowing down of fluctuations in an anatomically constrained dynamical system. Altogether, the reduction from spiking models to statistical moments presented here provides a new framework to explicitly understand the building up of FC through neuronal dynamics underpinned by anatomical connections and to drive hypotheses in task-evoked studies and for clinical applications.

Oligomerization and DNA binding of Ler, a master regulator of pathogenicity of enterohemorrhagic and enteropathogenic Escherichia coli

Ler is a DNA-binding, oligomerizable protein that regulates pathogenicity islands in enterohemorrhagic and enteropathogenic Escherichia coli strains. Ler counteracts the transcriptional silencing effect of H-NS, another oligomerizable nucleoid-associated protein. We studied the oligomerization of Ler in the absence and presence of DNA by atomic force microscopy. Ler forms compact particles with a multimodal size distribution corresponding to multiples of 35 units of Ler. DNA wraps around Ler particles that contain more than 1516 Ler monomers. The resulting shortening of the DNA contour length is in agreement with previous measurements of the length of DNA protected by Ler in footprinting assays. We propose that the repetition unit corresponds to the number of monomers per turn of a tight helical Ler oligomer. While the repressor (H-NS) and anti-repressor (Ler) have similar DNA-binding domains, their oligomerization domains are unrelated. We suggest that the different oligomerization behavior of the two proteins explains the opposite results of their interaction with the same or proximal regions of DNA.

Linear relations between writhe and minimal crossing number in Conway families of ideal knots and links

We showed earlier how to predict the writhe of any rational knot or link in its ideal geometric configuration, or equivalently the average of the 3D writhe over statistical ensembles of random configurations of a given knot or link (Cerf and Stasiak 2000 Proc. Natl Acad. Sci. USA 97 3795). There is no general relation between the minimal crossing number of a knot and the writhe of its ideal geometric configuration. However, within individual families of knots linear relations between minimal crossing number and writhe were observed (Katritch et al 1996 Nature 384 142). Here we present a method that allows us to express the writhe as a linear function of the minimal crossing number within Conway families of knots and links in their ideal configuration. The slope of the lines and the shift between any two lines with the same

Associations of serum uric acid and SLC2A9 variant with depressive and anxiety disorders: a population-based study

Background: Limited information exists regarding the association between serum uric acid (SUA) and psychiatric disorders. We explored the relationship between SUA and subtypes of major depressive disorder (MDD) and specific anxiety disorders. Additionally, we examined the association of SLC2A9 rs6855911 variant with anxiety disorders. Methods: We conducted a cross-sectional analysis on 3,716 individuals aged 35-66 years previously selected for the population-based CoLaus survey and who agreed to undergo further psychiatric evaluation. SUA was measured using uricase-PAP method. The French translation of the semi-structured Diagnostic Interview for Genetic Studies was used to establish lifetime and current diagnoses of depression and anxiety disorders according to the DSM-IV criteria. Results: Men reported significantly higher levels of SUA compared to women (357}74 μmol/L vs. 263}64 μmol/L). The prevalence of lifetime and current MDD was 44% and 18% respectively while the corresponding estimates for any anxiety disorders were 18% and 10% respectively. A quadratic hockey-stick shaped curve explained the relationship between SUA and social phobia better than a linear trend. However, with regards to the other specific anxiety disorders and other subtypes of MDD, there was no consistent pattern of association. Further analyses using SLC2A9 rs6855911 variant, known to be strongly associated with SUA, supported the quadratic relationship observed between SUA phenotype and social phobia. Conclusions: A quadratic relationship between SUA and social phobia was observed consistent with a protective effect of moderately elevated SUA on social phobia, which disappears at higher concentrations. Further studies are needed to confirm our observations.

The function of a new regulator of epidermal differentiation, HOP, and pathomechanisms underlying epidermolytic hyperkeratosis

The process of epidermal differentiation involves proliferation, differentiation, migration and maturation of keratinocytes to form an impermeable barrier against water loss and outside environment. It is controlled by highly balanced regulatory machinery, involving many molecules that are still under investigation.Homeobox proteins are involved in body patterning and morphogenesis of organs and are studied as potentially good candidates to regulate this process. In the first project we investigated the role of a protein named HOP which belongs to a group of homeobox proteins. Even if HOP is a small protein almost completely composed of the homeodomain and without DNA binding capacity, it is considered as transcriptional regulator in different tissues. HOP interacts with serum response factor (SRF) and histone deacetylase type 2 (HDAC2). By microarray analysis we found that HOP expression increases in cultured human primary keratinocytes (NHK) which undergo calcium-induced differentiation. HOP protein was localized in granular layer of the epidermis of healthy individuals. Lack of HOP was demonstrated in psoriatic lesions, whereas a strong expression was demonstrated in the lesional skin of patients affected with lichen planus (LP). Since LP is characterized by hypergranulosis while psoriatic lesions by progressive lack of the granular layer, the obtained data indicated that HOP might have a potential function in granular layer of epidermis. To investigate HOP function, we inhibited its expression by using HOP specific StealthRNAi and we overexpressed HOP using lentiviral vectors in differentiating NHK. The conclusion of both experiments indicated that HOP positively regulates the expression of late differentiation markers, such as profilaggrin, loricrin and transglutaminase 1. The in vitro data were next confirmed in vivo using HOP knockout mouse model.The second part of my study involved analysis of mechanisms underlying the pathogenesis of epidermolytic hyperkeratosis (EHK). EHK is a genetic disorder characterized by erythema, skin blistering, keratinocyte hyperproliferation and hyperkeratosis. EHK is caused by mutations in keratin 1 or 10 (K1, K10) which are major structural proteins of differentiated keratinocytes and participate in the cellular scaffold formation. To investigate how the structural proteins carrying mutations alter cellular signaling, we established an in vitro model for EHK by overexpression of one of the most common K10 mutations reported so far (K10R156H), in primary human keratinocytes. In order to mimic the in vivo situation, mutated keratinocytes growing on silicone membranes were subjected to mechanical stretch. We observed strong collapse of KIF in K10R156H keratinocytes when subjected to stretch for 30 minutes. Our data demonstrated stronger activation of p38, a member of MAPK stress signaling pathways, in K10R156H when compared to control cells. We demonstrated also that K10R156H keratinocytes showed an induction of TNF-α and RANTES release in response to stretch.Taken together these studies characterize a novel regulator of epidermal differentiation - HOP and demonstrate new aspects implicated in the pathogenesis of EHK.

The role of NFI-C liver regeneration and its potential function as a general regulator of regenerative processes

Collectively, research aimed to understand the regeneration of certain tissues has unveiled the existence of common key regulators. Knockout studies of the murine Nuclear Factor I-C (NFI-C) transcription factor revealed a misregulation of growth factor signaling, in particular that of transforming growth factor ß-1 (TGF-ßl), which led to alterations of skin wound healing and the growth of its appendages, suggesting it may be a general regulator of regenerative processes. We sought to investigate this further by determining whether NFI-C played a role in liver regeneration. Liver regeneration following two-thirds removal of the liver by partial hepatectomy (PH) is a well-established regenerative model whereby changes elicited in hepatocytes following injury lead to a rapid, phased proliferation. However, mechanisms controlling the action of liver proliferative factors such as transforming growth factor-ßl (TGF-ß1) and plasminogen activator inhibitor-1 (PAI-1) remain largely unknown. We show that the absence of NFI-C impaired hepatocyte proliferation due to an overexpression of PAI-1 and the subsequent suppression of urokinase plasminogen (uPA) activity and hepatocyte growth factor (HGF) signaling, a potent hepatocyte mitogen. This indicated that NFI-C first acts to promote hepatocyte proliferation at the onset of liver regeneration in wildtype mice. The subsequent transient down regulation of NFI-C, as can be explained by a self- regulatory feedback loop with TGF-ßl, may limit the number of hepatocytes entering the first wave of cell division and/or prevent late initiations of mitosis. Overall, we conclude that NFI-C acts as a regulator of the phased hepatocyte proliferation during liver regeneration. Taken together with NFI-C's actions in other in vivo models of (re)generation, it is plausible that NFI-C may be a general regulator of regenerative processes. - L'ensemble des recherches visant à comprendre la régénération de certains tissus a permis de mettre en évidence l'existence de régulateurs-clés communs. L'étude des souris, dépourvues du gène codant pour le facteur de transcription NFI-C (Nuclear Factor I-C), a montré des dérèglements dans la signalisation de certains facteurs croissance, en particulier du TGF-ßl (transforming growth factor-ßl), ce qui conduit à des altérations de la cicatrisation de la peau et de la croissance des poils et des dents chez ces souris, suggérant que NFI-C pourrait être un régulateur général du processus de régénération. Nous avons cherché à approfondir cette question en déterminant si NFI-C joue un rôle dans la régénération du foie. La régénération du foie, induite par une hépatectomie partielle correspondant à l'ablation des deux-tiers du foie, constitue un modèle de régénération bien établi dans lequel la lésion induite conduit à la prolifération rapide des hépatocytes de façon synchronisée. Cependant, les mécanismes contrôlant l'action de facteurs de prolifération du foie, comme le facteur de croissance TGF-ßl et l'inhibiteur de l'activateur du plasminogène PAI-1 (plasminogen activator inhibitor-1), restent encore très méconnus. Nous avons pu montrer que l'absence de NFI-C affecte la prolifération des hépatocytes, occasionnée par la surexpression de PAI-1 et par la subséquente suppression de l'activité de la protéine uPA (urokinase plasminogen) et de la signalisation du facteur de croissance des hépatocytes HGF (hepatocyte growth factor), un mitogène puissant des hépatocytes. Cela indique que NFI-C agit en premier lieu pour promouvoir la prolifération des hépatocytes au début de la régénération du foie chez les souris de type sauvage. La subséquente baisse transitoire de NFI-C, pouvant s'expliquer par une boucle rétroactive d'autorégulation avec le facteur TGF-ßl, pourrait limiter le nombre d'hépatocytes qui entrent dans la première vague de division cellulaire et/ou inhiber l'initiation de la mitose tardive. L'ensemble de ces résultats nous a permis de conclure que NFI-C agit comme un régulateur de la prolifération des hépatocytes synchrones au cours de la régénération du foie.