Expectation Propagation with Factorizing Distributions: A Gaussian Approximation and Performance Results for Simple Models

We discuss the expectation propagation (EP) algorithm for approximate Bayesian inference using a factorizing posterior approximation. For neural network models, we use a central limit theorem argument to make EP tractable when the number of parameters is large. For two types of models, we show that EP can achieve optimal generalization performance when data are drawn from a simple distribution.

Superior electric double layer capacitors using ordered mesoporous carbons

This paper reports for the first time superior electric double layer capacitive properties of ordered mesoporous carbon (OMCs) with varying ordered pore symmetries and mesopore structure. Compared to commercially used activated carbon electrode, Maxsorb, these OMC carbons have superior capacitive behavior, power output and high-frequency performance in EDLCs due to the unique structure of their mesopore network, which is more favorable for fast ionic transport than the pore networks in disordered microporous carbons. As evidenced by N-2 sorption, cyclic voltammetry and frequency response measurements, OMC carbons with large mesopores, and especially with 2-D pore symmetry, show superior capacitive behaviors (exhibiting a high capacitance of over 180 F/g even at very high sweep rate of 50 mV/s, as compared to much reduced capacitance of 73 F/g for Maxsorb at the same sweep rate). OMC carbons can provide much higher power density while still maintaining good energy density. OMC carbons demonstrate excellent high-frequency performances due to its higher surface area in pores larger than 3 nm. Such ordered mesoporous carbons (OMCs) offer a great potential in EDLC capacitors, particularly for applications where high power output and good high-frequency capacitive performances are required. (C) 2005 Elsevier Ltd. All rights reserved.

Deficits on self ordered tasks associated with hyperostosis frontalis interna

A 74 year old patient, EW, with dorsolateral frontal cortical compression due to hyperostosis frontalis interna, in the absence of the Morgagni or Stewart-Morel syndromes, is described. In addition to conventional neuropsychological measures EW was administered one nonspatial and two spatial self ordered working memory tasks, as well as a standard measure of fluid intelligence or g. She showed impaired performance on all three self ordered working memory tasks compared with a normal control group of 10 subjects matched for age, education, sex, and IQ. By contrast, her performance on the fluid intelligence test was comparable with that of the controls. It is concluded that the compression of dorsolateral frontal cortex accompanying hyperostosis frontalis interna may produce selective cognitive impairment.

Alterations in Cytokine Profile and Dendritic Cells Subsets in Peripheral Blood of Rheumatoid Arthritis Patients before and after Biologic Therapy

Rheumatoid arthritis (RA) is an autoimmune disorder characterized by chronic joint inflammation and continuous immune cell infiltration in the synovium. These changes are linked to inflammatory cytokine release, leading to eventual destruction of cartilage and bone. During the last decade new therapeutic modalities have improved the prognosis, with the introduction of novel biological response modifiers including anti-TNF alpha CTLA4Ig and, more recently, anti-IL6. In the present study we looked at the immunological effects of these three forms of therapy. Serum, obtained from patients with RA was analyzed for TNF alpha, IL6, IL10, IFN gamma, and VEGF, and in parallel, circulating plasmacytoid and myeloid dendritic cells (DC) were enumerated before and after three infusions of the respective biological treatments. After treatment with anti-IL6, we found a significant reduction of IL6 and TNF alpha levels and the percentage of both DC subsets decreased. Although the results did not reach statistical significance for anti-TNF alpha treatment, similar trends were observed. Meanwhile, CTLA4Ig therapy led to the reduction IFN gamma levels only. None of the treatments modified significantly VEGF or IL10 levels. These findings may explain why patients with RA improve more rapidly on IL-6 therapy than with the other two modalities.

Phenotypic characterization of five dendritic cell subsets in human tonsils

Heterogeneous expression of several antigens on the three currently defined tonsil dendritic cell (DC) subsets encouraged us to re-examine tonsil DCs using a new method that minimized DC differentiation and activation during their preparation. Three-color flow cytometry and dual-color immunohistology was used in conjunction with an extensive panel of antibodies to relevant DC-related antigens to analyze lin(-) HLA-DR+ tonsil DCs. Here we identify, quantify, and locate five tonsil DC subsets based on their relative expression of the HLA-DR, CD11c, CD13, and CD123 antigens. In situ localization identified four of these DC subsets as distinct interdigitating DC populations. These included three new interdigitating DC subsets defined as HLA-DRhi CD11c(+) DCs, HLA-DRmod CD11c+ CD13(+) DCs, and HLA-DRmod CD11c(-) CD123(-) DCs, as well as the plasmacytoid DCs (HLA-DRmod CD11c- CD123(+)). These subsets differed in their expression of DC-associated differentiation/activation antigens and co-stimulator molecules including CD83, CMRF-44, CMRF-56, 2-7, CD86, and 4-1BB ligand. The fifth HLA-DRmod CD11c(+) DC subset was identified as germinal center DCs, but contrary to previous reports they are redefined as lacking the CD13 antigen. The definition and extensive phenotypic analysis of these five DC subsets In human tonsil extends our understanding of the complexity of DC biology.

Functional heterogeneity within rhodamine123(lo) Hoechst33342(lo/sp) primitive hemopoietic stem cells revealed by pyronin Y

Objective. The aim of this study was to determine the function of primitive hematopoietic stem cells (PHSC) at phases G(0) and G(1) of the cell cycle. Materials and Methods. A combination of supravital dyes rhodamine123 (Rh), Hoechst33342 (Ho), and pyronin (PY) was used to isolate the G(0) and G(1) subsets of PHSC. A competitive repopulation assay was used to evaluate their in vivo function. Results. We confirmed that the Rh(lo)Lin(-)Kit(+)Sca-1(+) PHSC were relatively quiescent when compared with the more mature Rh(hi)Lin(-)Kit(+)Sca-1 HSC and Rh(hi)Lin(-)Kit(+)Sca-1(-) progenitors. In addition, cells with Rh(lo)Lin(-)Kit(+)Sca-1(+), Rh(lo)Ho(lo)Lin(-)Sca-1(+), or Rh(lo)Ho(sp)Lin(-)Sca-1(+) phenotypes identified the same cell population. We further subfractionated the Rh(lo)Ho(lo/sp)Lin(-)Sca-1(+) PHSC using PY into PYlo and PYhi subsets. Limiting dilution analysis revealed that the frequency of long-term in vivo competitive repopulating units (CRU) of the (PYRhHolo/sp)-Rh-lo-Ho-lo PHSC was 1 in 10 cells, whereas there was at least a three-fold lower frequency in those isolated at the G(1) phase (PYhi) We found a dose-dependent PY-mediated cytotoxicity that at moderate concentration affected most of the murine hematopoietic compartment but spared the early HSC compartment. Conclusion. Our data confirm that the HSC compartment is hierarchically ordered on the basis of quiescence and further extend this concept to PY-mediated cytotoxicity. PY supravital dye can be used to reveal functional heterogeneity within the (RhHolosp)-Ho-lo PHSC population but is of limited use in dissecting the relatively more mature hematopoietic stem/progenitor cell population. (C) 2001 International Society for Experimental Hematology. Published by Elsevier Science Inc.

Fitting genetic models to twin data with binary and ordered categorical responses: A comparison of structural equation modelling and Bayesian hierarchical models

We compare Bayesian methodology utilizing free-ware BUGS (Bayesian Inference Using Gibbs Sampling) with the traditional structural equation modelling approach based on another free-ware package, Mx. Dichotomous and ordinal (three category) twin data were simulated according to different additive genetic and common environment models for phenotypic variation. Practical issues are discussed in using Gibbs sampling as implemented by BUGS to fit subject-specific Bayesian generalized linear models, where the components of variation may be estimated directly. The simulation study (based on 2000 twin pairs) indicated that there is a consistent advantage in using the Bayesian method to detect a correct model under certain specifications of additive genetics and common environmental effects. For binary data, both methods had difficulty in detecting the correct model when the additive genetic effect was low (between 10 and 20%) or of moderate range (between 20 and 40%). Furthermore, neither method could adequately detect a correct model that included a modest common environmental effect (20%) even when the additive genetic effect was large (50%). Power was significantly improved with ordinal data for most scenarios, except for the case of low heritability under a true ACE model. We illustrate and compare both methods using data from 1239 twin pairs over the age of 50 years, who were registered with the Australian National Health and Medical Research Council Twin Registry (ATR) and presented symptoms associated with osteoarthritis occurring in joints of the hand.

Conductivity/microstructure study of a pyrochlore-type ordered phase in the compound Zr0.75Ce0.08Nd0.17O1.92

The compound Zr0.75Ce0.08Nd0.17O1.92 was investigated as part of a much larger electrical conductivity/microstructure study of the systems ZrO2-CeO2-M2O3 (where M=Nd, Sm, ..., Yb) [Solid State Ionics (2002)]. Electrical conductivity measurements performed in air at 800 degreesC showed significant conductivity degradation over a period of 200 h. Investigation of the annealed and as-fired specimens by ATEM revealed the presence of an emerging, ordered pyrochlore-type phase within the Zr0.75Ce0.08Nd0.17O1.92 defect-fluorite solid solution at much lower dopant levels than observed previously for zirconia binary systems. (C) 2002 Elsevier Science B.V. All rights reserved.

An EM-based Semi-Parametric Mixture Model Approach to the Regression Analysis of Competing-Risks Data

We consider a mixture model approach to the regression analysis of competing-risks data. Attention is focused on inference concerning the effects of factors on both the probability of occurrence and the hazard rate conditional on each of the failure types. These two quantities are specified in the mixture model using the logistic model and the proportional hazards model, respectively. We propose a semi-parametric mixture method to estimate the logistic and regression coefficients jointly, whereby the component-baseline hazard functions are completely unspecified. Estimation is based on maximum likelihood on the basis of the full likelihood, implemented via an expectation-conditional maximization (ECM) algorithm. Simulation studies are performed to compare the performance of the proposed semi-parametric method with a fully parametric mixture approach. The results show that when the component-baseline hazard is monotonic increasing, the semi-parametric and fully parametric mixture approaches are comparable for mildly and moderately censored samples. When the component-baseline hazard is not monotonic increasing, the semi-parametric method consistently provides less biased estimates than a fully parametric approach and is comparable in efficiency in the estimation of the parameters for all levels of censoring. The methods are illustrated using a real data set of prostate cancer patients treated with different dosages of the drug diethylstilbestrol. Copyright (C) 2003 John Wiley Sons, Ltd.

Multiple ordered phases in Langmuir-Blodgett films of cadmium arachidate at elevated temperatures

Using synchrotron X-ray grazing incidence diffraction, superlattice structures have been observed to develop in Langmuir-Blodgett films of cadmium arachidate as the temperature is raised. The previously reported superstructure in the stacked lamellae at room temperature changes at about 70 degreesC and there are further changes at about 90 and 103 degreesC before the major phase transition from stacked lamellae to hexagonally packed rods occurs at 107 degreesC (Langmuir 1997, 13, 1602). Between 70 and 103 degreesC there is a 1 x 10 one-dimensional in-plane superstructure, which is commensurate with the local structure and has an interlayer shift along [01] by a distance of b (of the local structure) at lower temperatures, and a further shift at about 90 degreesC. At lower (