Massimo Bilancia

KIR2DL3 and the KIR ligand groups HLA-A-Bw4 and HLA-C2 predict the outcome of hepatitis B virus infection

Killer immunoglobulin‐like receptors (KIRs) regulate the activation of natural killer cells through their interaction with human leucocyte antigens (HLA). KIR and HLA loci are highly polymorphic, and certain HLA‐KIR combinations have been found to protect against viral infections. In this study, we analysed whether the KIR/HLA repertoire may influence the course of hepatitis B virus (HBV) infection. Fifty‐seven subjects with chronic hepatitis B (CHB), 44 subjects with resolved HBV infection and 60 healthy uninfected controls (HC) were genotyped for KIR and their HLA ligands. The frequency of the HLA‐A‐Bw4 ligand group was higher in CHB (58%) than subjects with resolved infection (23%) (crude OR, 4.67; P<.001) and HC (10%) (crude OR, 12.38; P<.001). Similar results were obtained for the HLA‐C2 ligand group, more frequent in CHB (84%), than subjects with resolved infection (70%) (crude OR, 2.24; P<.10) and HC (60%) (crude OR, 3.56; P<.01). Conversely, the frequency of KIR2DL3 was lower in CHB (81%) than in subjects with resolved infection (98%) (crude OR, 0.10; P<.05). These results suggest a detrimental role of HLA‐A‐Bw4 and HLA‐C2 groups, which are associated with the development of CHB, and a protective role of KIR2DL3. A stepwise variable selection procedure, based on multiple logistic regression analysis, identified these three predictive variables as the most relevant, featuring high specificity (90.9%) and positive predictive value (87.5%) for the development of CHB. Our results suggest that a combination of KIR/HLA gene/alleles is able to predict the outcome of HBV infection.

Relationships between mild PM10 and ozone urban air levels and spontaneous abortion: clues for primary prevention.

The effects of environmental pollution on spontaneous abortion (SAB) are still unclear. Records of SAB were collected from five cities (514,996 residents) and correlated with PM10, NO2 and ozone levels. Median pollutant concentrations were below legal limits. Monthly SABs positively correlated with PM10 and ozone levels but not with NO2 levels. The mean monthly SAB rate increase was estimated equal to 19.7 and 33.6 % per 10 μg/m3 increase in PM10 or ozone concentration, respectively. Higher values of PM10 and SABs were evident in cities with- than in those without pollutant industries, with a number of SABs twofolds higher in the former group. In conclusion, SAB occurrence is affected by PM10 (particularly if industrial areas are present) and ozone concentrations, also at levels below the legal limits. Thus, SAB might be considered, at least in part, a preventable condition.

Anthropogenic CO 2 Emissions and Global Warming: Evidence from Granger Causality Analysis

This note reports an updated analysis of global climate change and its relationship with Carbon Dioxide (CO2) emissions: advanced methods rooted in econometrics are applied to bivariate climatic time series. We found a strong evidence for the absence of Granger causality from CO2 emissions to global surface temperature: we can conclude that our findings point out that the hypothesis of anthropogenically-induced climate change still need a conclusive confirmation using the most appropriate methods for data analysis.

A Model-Based Scan Statistics for Detecting Geographical Clustering of Disease

The classical likelihood ratio spatial scan statistics has been widely used in spatial epidemiology for disease cluster detection. The question is whether the geographic incidence pattern is due to random fluctuations or the map reflects true underlying geographical variation due to etiologic risk factors. The hypothesis underlying the classic scan statistics assume that disease counts in different locations have independent Poisson distribution; unfortunately, outcomes in spatial units are often not independent of each other. Risk estimates of areas that are close to each other will tend to be positively correlated as they share a number of spatially varying characteristics. Ignoring the overdispersion caused by spatial autocorrelation leads to incorrect results. To overcome this difficulty, we propose a model-based approach adjusting for area-specific fixed-effects measuring potential effect modifiers, and for large-scale geographical variation of etiologic factors that vary continuously in space and are not expressly present within the model. We apply our methodology to the spatial distribution of lung cancer male mortality occurred in the province of Lecce, Italy, during the period 1992-2001.

Timescale effect estimation in time-series studies of air pollution and health: A Singular Spectrum Analysis approach

A wealth of epidemiological data suggests an association between mortality/morbidity from pulmonary and cardiovascular adverse events and air pollution, but uncertainty remains as to the extent implied by those associations although the abundance of the data. In this paper we describe an SSA (Singular Spectrum Analysis) based approach in order to decompose the time-series of particulate matter concentration into a set of exposure variables, each one representing a different timescale. We implement our methodology to investigate both acute and long-term effects of

Predicting Bug-Fix Time: Using Standard Versus Topic-Based Text Categorization Techniques

PM_{10}

Bayesian scanning of spatial disease rates with integrated nested Laplace approximation (INLA)

exposure on morbidity from respiratory causes within the urban area of Bari, Italy.

Geographical Disparities in Mortality Rates: Spatial Data Mining and Bayesian Hierarchical Modeling

In modern software development, finding and fixing bugs is a vital part of software development and quality assurance. Once a bug is reported, it is typically recorded in the Bug Tracking System, and is assigned to a developer to resolve (bug triage). Current practice of bug triage is largely a manual collaborative process, which is often time-consuming and error-prone. Predicting on the basis of past data the time to fix a newly-reported bug has been shown to be an important target to support the whole triage process. Many researchers have, therefore, proposed methods for automated bug-fix time prediction, largely based on statistical prediction models exploiting the attributes of bug reports. However, existing algorithms often fail to validate on multiple large projects widely-used in bug studies, mostly as a consequence of inappropriate attribute selection [2]. In this paper, instead of focusing on attribute subset selection, we explore an alternative promising approach consisting of using all available textual information. The problem of bug-fix time estimation is then mapped to a text categorization problem. We consider a multi-topic Supervised Latent Dirichlet Allocation (SLDA) model, which adds to Latent Dirichlet Allocation a response variable consisting of an unordered binary target variable, denoting time to resolution discretized into FAST (negative class) and SLOW (positive class) labels. We have evaluated SLDA on four large-scale open source projects. We show that the proposed model greatly improves recall, when compared to standard single topic algorithms.

Economic Evaluation and Statistical Methods for Detecting Hot Spots of Social and Housing Difficulties in Urban Policies

Among the many tools suited to detect local clusters in group-level data, Kulldorff–Nagarwalla’s spatial scan statistic gained wide popularity (Kulldorff and Nagarwalla in Stat Med 14(8):799–810, 1995). The underlying assumptions needed for making statistical inference feasible are quite strong, as counts in spatial units are assumed to be independent Poisson distributed random variables. Unfortunately, outcomes in spatial units are often not independent of each other, and risk estimates of areas that are close to each other will tend to be positively correlated as they share a number of spatially varying characteristics. We therefore introduce a Bayesian model-based algorithm for cluster detection in the presence of spatially autocorrelated relative risks. Our approach has been made possible by the recent development of new numerical methods based on integrated nested Laplace approximation, by which we can directly compute very accurate approximations of posterior marginals within short computational time (Rue et al. in JRSS B 71(2):319–392, 2009). Simulated data and a case study show that the performance of our method is at least comparable to that of Kulldorff–Nagarwalla’s statistic.

Predicting Student Dropouts in Higher Education Using Supervised Classification Algorithms

Achieving health equity has been identified as a major international challenge since the 1978 declaration of Alma Ata. Disease risk maps provide important clues concerning many aspects of health equity, such as etiology risk factors involved by occupational and environmental exposures, as well as gender-related and socioeconomic inequalities. This explains why epidemiological disease investigation should always include an assessment of the spatial variation of disease risk, with the objective of producing a representation of important spatial effects while removing any noise. Bearing in mind this goal, this review covers basic and more advanced aspects of Bayesian models for disease mapping, and methods to analyze whether the spatial distribution of the disease risk closely follows that of underlying population at risk, or there exist some nonrandom local patterns (disease clusters) which may suggest a further explanation for disease etiology. We provide a practical illustration by analyzing the spatial distribution of liver cancer mortality in Apulia, Italy, during the 2000–2005 quinquennial. (Massimo Bilancia wrote Sects. 1.1.2, 1.1.4, 1.1.6, 1.2.1, 1.2.3, 1.2.5. Giusi Graziano wrote Sects. 1.1.1, 1.1.3, 1.1.5, 1.2.2, 1.2.4, 1.2.6. Giacomo Demarinis wrote the software for data analysis. Section 1.3 was written jointly. The three authors read and approved the final manuscript. We wish to thank Maria Rosa Debellis, Department of Neuroscience and Sense Organs, University of Bari, Italy, and Claudia Monte PhD, Department of Physics, University of Bari, Italy, for their valuable support.)