Bayesian parameter estimation for latent Markov random fields and social networks

Undirected graphical models are widely used in statistics, physics and machine vision. However Bayesian parameter estimation for undirected models is extremely challenging, since evaluation of the posterior typically involves the calculation of an intractable normalising constant. This problem has received much attention, but very little of this has focussed on the important practical case where the data consists of noisy or incomplete observations of the underlying hidden structure. This paper specifically addresses this problem, comparing two alternative methodologies. In the first of these approaches particle Markov chain Monte Carlo (Andrieu et al., 2010) is used to efficiently explore the parameter space, combined with the exchange algorithm (Murray et al., 2006) for avoiding the calculation of the intractable normalising constant (a proof showing that this combination targets the correct distribution in found in a supplementary appendix online). This approach is compared with approximate Bayesian computation (Pritchard et al., 1999). Applications to estimating the parameters of Ising models and exponential random graphs from noisy data are presented. Each algorithm used in the paper targets an approximation to the true posterior due to the use of MCMC to simulate from the latent graphical model, in lieu of being able to do this exactly in general. The supplementary appendix also describes the nature of the resulting approximation.

Heme-induced biomarkers associated with red meat promotion of colon cancer are not modulated by the intake of nitrite

Red and processed meat consumption is associated with the risk of colorectal cancer. Three hypotheses are proposed to explain this association, via heme-induced oxidation of fat, heterocyclic amines, or N-nitroso compounds. Rats have often been used to study these hypotheses, but the lack of enterosalivary cycle of nitrate in rats casts doubt on the relevance of this animal model to predict nitroso- and heme-associated human colon carcinogenesis. The present study was thus designed to clarify whether a nitrite intake that mimics the enterosalivary cycle can modulate hemeinduced nitrosation and fat peroxidation. This study shows that, in contrast with the starting hypothesis, drinking water added with nitrite to mimic the salivary nitrite content did not change the effect of hemoglobin on biochemicalmarkers linked to colon carcinogenesis, notably lipid peroxidation and cytotoxic activity in the colon of rat. However, ingested sodium nitrite increased fecal nitrosocompounds level, but their fecal concentration and their nature (iron-nitrosyl) would probably not be associated with an increased risk of cancer.We thus suggest that the rat model could be relevant for study the effect of red meat on colon carcinogenesis, in spite of the lack of nitrite in the saliva of rats.

The linearly modulated IRSL red emission from feldspars

This paper uses the linear modulation technique to study red IRSL emission of potassium feldspars. Sub-samples were subjected to various pre-treatment and measurement conditions in an attempt to understand the relevant mechanisms of charge transfer. The linear modulation curves fitted most successfully to a sum of three first order components and we present supporting empirical evidence for the presence of three separate signal components. Additionally, the form of the red emission was observed to closely resemble the UV emission, implying the same donor charge concentrations may supply different recombination centres (assuming emission wavelength depends on centre type).

Business cycle asymmetries: characterisation and testing based on Markov-switching autoregressions

Tests for business cycle asymmetries are developed for Markov-switching autoregressive models. The tests of deepness, steepness, and sharpness are Wald statistics, which have standard asymptotics. For the standard two-regime model of expansions and contractions, deepness is shown to imply sharpness (and vice versa), whereas the process is always nonsteep. Two and three-state models of U.S. GNP growth are used to illustrate the approach, along with models of U.S. investment and consumption growth. The robustness of the tests to model misspecification, and the effects of regime-dependent heteroscedasticity, are investigated.

Threshold autoregressive and Markov switching models: an application to commercial real estate

Although financial theory rests heavily upon the assumption that asset returns are normally distributed, value indices of commercial real estate display significant departures from normality. In this paper, we apply and compare the properties of two recently proposed regime switching models for value indices of commercial real estate in the US and the UK, both of which relax the assumption that observations are drawn from a single distribution with constant mean and variance. Statistical tests of the models' specification indicate that the Markov switching model is better able to capture the non-stationary features of the data than the threshold autoregressive model, although both represent superior descriptions of the data than the models that allow for only one state. Our results have several implications for theoretical models and empirical research in finance.

Noisy Monte Carlo: convergence of Markov chains with approximate transition kernels

Monte Carlo algorithms often aim to draw from a distribution π by simulating a Markov chain with transition kernel P such that π is invariant under P. However, there are many situations for which it is impractical or impossible to draw from the transition kernel P. For instance, this is the case with massive datasets, where is it prohibitively expensive to calculate the likelihood and is also the case for intractable likelihood models arising from, for example, Gibbs random fields, such as those found in spatial statistics and network analysis. A natural approach in these cases is to replace P by an approximation Pˆ. Using theory from the stability of Markov chains we explore a variety of situations where it is possible to quantify how ’close’ the chain given by the transition kernel Pˆ is to the chain given by P . We apply these results to several examples from spatial statistics and network analysis.

Hidden Markov analysis describes dive patterns in semiaquatic animals

Existing methods of dive analysis, developed for fully aquatic animals, tend to focus on frequency of behaviors rather than transitions between them. They, therefore, do not account for the variability of behavior of semiaquatic animals, and the switching between terrestrial and aquatic environments. This is the first study to use hidden Markov models (HMM) to divide dives of a semiaquatic animal into clusters and thus identify the environmental predictors of transition between behavioral modes. We used 18 existing data sets of the dives of 14 American mink (Neovison vison) fitted with time-depth recorders in lowland England. Using HMM, we identified 3 behavioral states (1, temporal cluster of dives; 2, more loosely aggregated diving within aquatic activity; and 3, terminal dive of a cluster or a single, isolated dive). Based on the higher than expected proportion of dives in State 1, we conclude that mink tend to dive in clusters. We found no relationship between temperature and the proportion of dives in each state or between temperature and the rate of transition between states, meaning that in our study area, mink are apparently not adopting different diving strategies at different temperatures. Transition analysis between states has shown that there is no correlation between ambient temperature and the likelihood of mink switching from one state to another, that is, changing foraging modes. The variables provided good discrimination and grouped into consistent states well, indicating promise for further application of HMM and other state transition analyses in studies of semiaquatic animals.

Transition from suppressed to active convection modulated by a weak-temperature gradient derived large-scale circulation

Numerical simulations are performed to assess the influence of the large-scale circulation on the transition from suppressed to active convection. As a model tool, we used a coupled-column model. It consists of two cloud-resolving models which are fully coupled via a large-scale circulation which is derived from the requirement that the instantaneous domain-mean potential temperature profiles of the two columns remain close to each other. This is known as the weak-temperature gradient approach. The simulations of the transition are initialized from coupled-column simulations over non-uniform surface forcing and the transition is forced within the dry column by changing the local and/or remote surface forcings to uniform surface forcing across the columns. As the strength of the circulation is reduced to zero, moisture is recharged into the dry column and a transition to active convection occurs once the column is sufficiently moistened to sustain deep convection. Direct effects of changing surface forcing occur over the first few days only. Afterward, it is the evolution of the large-scale circulation which systematically modulates the transition. Its contributions are approximately equally divided between the heating and moistening effects. A transition time is defined to summarize the evolution from suppressed to active convection. It is the time when the rain rate within the dry column is halfway to the mean value obtained at equilibrium over uniform surface forcing. The transition time is around twice as long for a transition that is forced remotely compared to a transition that is forced locally. Simulations in which both local and remote surface forcings are changed produce intermediate transition times.

Decomposition of surface electromyographic signal using Hidden Markov Model

The detection of physiological signals from the motor system (electromyographic signals) is being utilized in the practice clinic to guide the therapist in a more precise and accurate diagnosis of motor disorders. In this context, the process of decomposition of EMG (electromyographic) signals that includes the identification and classification of MUAP (Motor Unit Action Potential) of a EMG signal, is very important to help the therapist in the evaluation of motor disorders. The EMG decomposition is a complex task due to EMG features depend on the electrode type (needle or surface), its placement related to the muscle, the contraction level and the health of the Neuromuscular System. To date, the majority of researches on EMG decomposition utilize EMG signals acquired by needle electrodes, due to their advantages in processing this type of signal. However, relatively few researches have been conducted using surface EMG signals. Thus, this article aims to contribute to the clinical practice by presenting a technique that permit the decomposition of surface EMG signal via the use of Hidden Markov Models. This process is supported by the use of differential evolution and spectral clustering techniques. The developed system presented coherent results in: (1) identification of the number of Motor Units actives in the EMG signal; (2) presentation of the morphological patterns of MUAPs in the EMG signal; (3) identification of the firing sequence of the Motor Units. The model proposed in this work is an advance in the research area of decomposition of surface EMG signals.

The uniqueness of signature problem in the non-Markov setting

We establish a general framework for a class of multidimensional stochastic processes over [0,1] under which with probability one, the signature (the collection of iterated path integrals in the sense of rough paths) is well-defined and determines the sample paths of the process up to reparametrization. In particular, by using the Malliavin calculus we show that our method applies to a class of Gaussian processes including fractional Brownian motion with Hurst parameter H>1/4, the Ornstein–Uhlenbeck process and the Brownian bridge.

Spontaneous facial mimicry is modulated by joint attention and autistic traits

Joint attention (JA) and spontaneous facial mimicry (SFM) are fundamental processes in social interactions, and they are closely related to empathic abilities. When tested independently, both of these processes have been usually observed to be atypical in individuals with autism spectrum conditions (ASC). However, it is not known how these processes interact with each other in relation to autistic traits. This study addresses this question by testing the impact of JA on SFM of happy faces using a truly interactive paradigm. Sixty-two neurotypical participants engaged in gaze-based social interaction with an anthropomorphic, gaze-contingent virtual agent. The agent either established JA by initiating eye contact or looked away, before looking at an object and expressing happiness or disgust. Eye tracking was used to make the agent's gaze behavior and facial actions contingent to the participants' gaze. SFM of happy expressions was measured by Electromyography (EMG) recording over the Zygomaticus Major muscle. Results showed that JA augments SFM in individuals with low compared with high autistic traits. These findings are in line with reports of reduced impact of JA on action imitation in individuals with ASC. Moreover, they suggest that investigating atypical interactions between empathic processes, instead of testing these processes individually, might be crucial to understanding the nature of social deficits in autism

Expression of the receptor tyrosine kinase ephb2 on dendritic cells is modulated by toll-like receptor ligation but is not required for t cell activation

The Eph receptor tyrosine kinases interact with their ephrin ligands on adjacent cells to facilitate contact-dependent cell communication. Ephrin B ligands are expressed on T cells and have been suggested to act as co-stimulatory molecules during T cell activation. There are no detailed reports of the expression and modulation of EphB receptors on dendritic cells, the main antigen presenting cells that interact with T cells. Here we show that mouse splenic dendritic cells (DC) and bone-marrow derived DCs (BMDC) express EphB2, a member of the EphB family. EphB2 expression is modulated by ligation of TLR4 and TLR9 and also by interaction with ephrin B ligands. Co-localization of EphB2 with MHC-II is also consistent with a potential role in T cell activation. However, BMDCs derived from EphB2 deficient mice were able to present antigen in the context of MHC-II and produce T cell activating cytokines to the same extent as intact DCs. Collectively our data suggest that EphB2 may contribute to DC responses, but that EphB2 is not required for T cell activation. This result may have arisen because DCs express other members of the EphB receptor family, EphB3, EphB4 and EphB6, all of which can interact with ephrin B ligands, or because EphB2 may be playing a role in another aspect of DC biology such as migration.

Lightning as a space-weather hazard: UK thunderstorm activity modulated by the passage of the heliospheric current sheet

Lightning flash rates, RL, are modulated by corotating interaction regions (CIRs) and the polarity of the heliospheric magnetic field (HMF) in near-Earth space. As the HMF polarity reverses at the heliospheric current sheet (HCS), typically within a CIR, these phenomena are likely related. In this study, RL is found to be significantly enhanced at the HCS and at 27 days prior/after. The strength of the enhancement depends on the polarity of the HMF reversal at the HCS. Near-Earth solar and galactic energetic particle fluxes are also ordered by HMF polarity, though the variations qualitatively differ from RL, with the main increase occurring prior to the HCS crossing. Thus, the CIR effect on lightning is either the result of compression/amplification of the HMF (and its subsequent interaction with the terrestrial system) or that energetic particle preconditioning of the Earth system prior to the HMF polarity change is central to solar wind lightning coupling mechanism.

Matrix Metalloproteinase Expression in Teeth with Apical Periodontitis Is Differentially Modulated by the Modality of Root Canal Treatment

Introduction: The objective of this study was to investigate the expression of matrix metalloproteinases (MM Ps) in apical periodontitis and during the periapical healing phase after root canal treatment. Methods: Apical periodontitis was induced in dog teeth, and root canal treatment was performed in a single visit or by using an additional calcium hydroxide root canal dressing. One hundred eighty days after treatment the presence of inflammation was examined, and tissues were stained to detect bacteria. Bacterial status was correlated to the degree of tissue organization, and to further investigate molecules involved in this process, tissues were stained for MMP-1, MMP-2, MMP-8, and MMP-9. Data were analyzed by using one-way analysis of variance followed by Tukey test or Kruskal-Wallis followed by Dunn test. Results: Teeth with apical periodontitis that had root canal therapy performed in a single visit presented an intense inflammatory cell infiltrate. Periapical tissue was extremely disorganized, and this was correlated with the presence of bacteria. Higher MMP expression was evident, similar to teeth with untreated apical periodontitis. In contrast, teeth with apical periodontitis submitted to root canal treatment with calcium hydroxide presented a lower inflammatory cell infiltrate. This group had moderately organized connective tissue, lower prevalence of bacteria, and lower number of MMP-positive cells, similar to healthy teeth submitted to treatment. Conclusions: Teeth treated with calcium hydroxide root canal dressing exhibited a lower percentage of bacterial contamination, a lower MMP expression, and a more organized extracellular matrix, unlike those treated in a single visit. This suggests that calcium hydroxide might be beneficial in tissue repair processes. (J Endod 2010;36:231-237)

Frequency of extreme rainfall events in Southern Brazil modulated by interannual and interdecadal variability

The frequency of extreme rainfall events in Southern Brazil is impacted by Ell Nino - Southern Oscillation (ENSO) episodes, especially in austral spring. There are two areas in which this impact is more significant: one is on the coast, where extreme events are more frequent during El Nino (EN) and the other one extends inland, where extreme events increase during EN and decrease during La Nina (LN). Atmospheric circulation patterns associated with severe rainfall in those areas are similar (opposite) to anomalous patterns characteristic of EN (LN) episodes, indicating why increase (decrease) of extreme events in EN (LN) episodes is favoured. The most recurrent precipitation patterns during extreme rainfall events in each of these areas are disclosed by Principal Component Analysis (PCA) and evidence the separation between extreme events in these areas: a severe precipitation event generally does not occur simultaneously in the coast and inland, although they may Occur inland and in the coastal region in sequence. Although EN predominantly enhances extreme rainfall, there are EN years in which fewer severe events occur than the average of neutral years, and also the enhancement of extreme rainfall is not uniform for different EN episodes, because the interdecadal non-ENSO variability also modulates significantly the frequency of extreme events in Southern Brazil. The inland region, which is more affected, shows increase (decrease) of extreme rainfall in association with the negative (positive) phase of the Atlantic Multidecadal Variability, with the negative (positive) phase of the Pacific Multidecadal Variability and with the positive (negative) phase of the Pacific Interdecadal Variability. Copyright (C) 2008 Royal Meteorological Society