ON MARGINALIZED MULTILEVEL MODELS AND THEIR COMPUTATION

Clustered data analysis is characterized by the need to describe both systematic variation in a mean model and cluster-dependent random variation in an association model. Marginalized multilevel models embrace the robustness and interpretations of a marginal mean model, while retaining the likelihood inference capabilities and flexible dependence structures of a conditional association model. Although there has been increasing recognition of the attractiveness of marginalized multilevel models, there has been a gap in their practical application arising from a lack of readily available estimation procedures. We extend the marginalized multilevel model to allow for nonlinear functions in both the mean and association aspects. We then formulate marginal models through conditional specifications to facilitate estimation with mixed model computational solutions already in place. We illustrate this approach on a cerebrovascular deficiency crossover trial.

RANDOM EFFECTS MODELS IN A META-ANALYSIS OF THE ACCURACY OF DIAGNOSTIC TESTS WITHIN A GOLD STANDARD IN THE PRESENCE OF MISSING DATA

In evaluating the accuracy of diagnosis tests, it is common to apply two imperfect tests jointly or sequentially to a study population. In a recent meta-analysis of the accuracy of microsatellite instability testing (MSI) and traditional mutation analysis (MUT) in predicting germline mutations of the mismatch repair (MMR) genes, a Bayesian approach (Chen, Watson, and Parmigiani 2005) was proposed to handle missing data resulting from partial testing and the lack of a gold standard. In this paper, we demonstrate an improved estimation of the sensitivities and specificities of MSI and MUT by using a nonlinear mixed model and a Bayesian hierarchical model, both of which account for the heterogeneity across studies through study-specific random effects. The methods can be used to estimate the accuracy of two imperfect diagnostic tests in other meta-analyses when the prevalence of disease, the sensitivities and/or the specificities of diagnostic tests are heterogeneous among studies. Furthermore, simulation studies have demonstrated the importance of carefully selecting appropriate random effects on the estimation of diagnostic accuracy measurements in this scenario.

Multilevel Latent Class Models with Dirichlet Mixing Distribution

Latent class analysis (LCA) and latent class regression (LCR) are widely used for modeling multivariate categorical outcomes in social sciences and biomedical studies. Standard analyses assume data of different respondents to be mutually independent, excluding application of the methods to familial and other designs in which participants are clustered. In this paper, we develop multilevel latent class model, in which subpopulation mixing probabilities are treated as random effects that vary among clusters according to a common Dirichlet distribution. We apply the Expectation-Maximization (EM) algorithm for model fitting by maximum likelihood (ML). This approach works well, but is computationally intensive when either the number of classes or the cluster size is large. We propose a maximum pairwise likelihood (MPL) approach via a modified EM algorithm for this case. We also show that a simple latent class analysis, combined with robust standard errors, provides another consistent, robust, but less efficient inferential procedure. Simulation studies suggest that the three methods work well in finite samples, and that the MPL estimates often enjoy comparable precision as the ML estimates. We apply our methods to the analysis of comorbid symptoms in the Obsessive Compulsive Disorder study. Our models' random effects structure has more straightforward interpretation than those of competing methods, thus should usefully augment tools available for latent class analysis of multilevel data.

A unification of models for meta-analysis of diagnostic accuracy studies

Studies of diagnostic accuracy require more sophisticated methods for their meta-analysis than studies of therapeutic interventions. A number of different, and apparently divergent, methods for meta-analysis of diagnostic studies have been proposed, including two alternative approaches that are statistically rigorous and allow for between-study variability: the hierarchical summary receiver operating characteristic (ROC) model (Rutter and Gatsonis, 2001) and bivariate random-effects meta-analysis (van Houwelingen and others, 1993), (van Houwelingen and others, 2002), (Reitsma and others, 2005). We show that these two models are very closely related, and define the circumstances in which they are identical. We discuss the different forms of summary model output suggested by the two approaches, including summary ROC curves, summary points, confidence regions, and prediction regions.

Assessment of use of microsatellite polymorphism analysis for improving spatial distribution tracking of echinococcus multilocularis

Alveolar echinococcosis (AE)--caused by the cestode Echinococcus multilocularis--is a severe zoonotic disease found in temperate and arctic regions of the northern hemisphere. Even though the transmission patterns observed in different geographical areas are heterogeneous, the nuclear and mitochondrial targets usually used for the genotyping of E. multilocularis have shown only a marked genetic homogeneity in this species. We used microsatellite sequences, because of their high typing resolution, to explore the genetic diversity of E. multilocularis. Four microsatellite targets (EmsJ, EmsK, and EmsB, which were designed in our laboratory, and NAK1, selected from the literature) were tested on a panel of 76 E. multilocularis samples (larval and adult stages) obtained from Alaska, Canada, Europe, and Asia. Genetic diversity for each target was assessed by size polymorphism analysis. With the EmsJ and EmsK targets, two alleles were found for each locus, yielding two and three genotypes, respectively, discriminating European isolates from the other groups. With NAK1, five alleles were found, yielding seven genotypes, including those specific to Tibetan and Alaskan isolates. The EmsB target, a tandem repeated multilocus microsatellite, found 17 alleles showing a complex pattern. Hierarchical clustering analyses were performed with the EmsB findings, and 29 genotypes were identified. Due to its higher genetic polymorphism, EmsB exhibited a higher discriminatory power than the other targets. The complex EmsB pattern was able to discriminate isolates on a regional and sectoral level, while avoiding overdistinction. EmsB will be used to assess the putative emergence of E. multilocularis in Europe.

Analysis of atmospheric transport and the effects of seasonal and interannual transport variability on measurements made at the Pico Mountain observatory

A non-hierarchical K-means algorithm is used to cluster 47 years (1960–2006) of 10-day HYSPLIT backward trajectories to the Pico Mountain (PM) observatory on a seasonal basis. The resulting cluster centers identify the major transport pathways and collectively comprise a long-term climatology of transport to the observatory. The transport climatology improves our ability to interpret the observations made there and our understanding of pollution source regions to the station and the central North Atlantic region. I determine which pathways dominate transport to the observatory and examine the impacts of these transport patterns on the O3, NOy, NOx, and CO measurements made there during 2001–2006. Transport from the U.S., Canada, and the Atlantic most frequently reaches the station, but Europe, east Africa, and the Pacific can also contribute significantly depending on the season. Transport from Canada was correlated with the North Atlantic Oscillation (NAO) in spring and winter, and transport from the Pacific was uncorrelated with the NAO. The highest CO and O3 are observed during spring. Summer is also characterized by high CO and O3 and the highest NOy and NOx of any season. Previous studies at the station attributed the summer time high CO and O3 to transport of boreal wildfire emissions (for 2002–2004), and boreal fires continued to affect the station during 2005 and 2006. The particle dispersion model FLEXPART was used to calculate anthropogenic and biomass-burning CO tracer values at the station in an attempt to identify the regions responsible for the high CO and O3 observations during spring and biomass-burning impacts in summer.

Long-term immunologic response to antiretroviral therapy in low-income countries: a collaborative analysis of prospective studies

BACKGROUND: Few data are available on the long-term immunologic response to antiretroviral therapy (ART) in resource-limited settings, where ART is being rapidly scaled up using a public health approach, with a limited repertoire of drugs. OBJECTIVES: To describe immunologic response to ART among ART patients in a network of cohorts from sub-Saharan Africa, Latin America, and Asia. STUDY POPULATION/METHODS: Treatment-naive patients aged 15 and older from 27 treatment programs were eligible. Multilevel, linear mixed models were used to assess associations between predictor variables and CD4 cell count trajectories following ART initiation. RESULTS: Of 29 175 patients initiating ART, 8933 (31%) were excluded due to insufficient follow-up time and early lost to follow-up or death. The remaining 19 967 patients contributed 39 200 person-years on ART and 71 067 CD4 cell count measurements. The median baseline CD4 cell count was 114 cells/microl, with 35% having less than 100 cells/microl. Substantial intersite variation in baseline CD4 cell count was observed (range 61-181 cells/microl). Women had higher median baseline CD4 cell counts than men (121 vs. 104 cells/microl). The median CD4 cell count increased from 114 cells/microl at ART initiation to 230 [interquartile range (IQR) 144-338] at 6 months, 263 (IQR 175-376) at 1 year, 336 (IQR 224-472) at 2 years, 372 (IQR 242-537) at 3 years, 377 (IQR 221-561) at 4 years, and 395 (IQR 240-592) at 5 years. In multivariable models, baseline CD4 cell count was the most important determinant of subsequent CD4 cell count trajectories. CONCLUSION: These data demonstrate robust and sustained CD4 response to ART among patients remaining on therapy. Public health and programmatic interventions leading to earlier HIV diagnosis and initiation of ART could substantially improve patient outcomes in resource-limited settings.

IDENTIFICATION OF GENES CONTROLLING BIOLOGICAL PROCESSES AND PATHWAYS THROUGH STATISTICAL ANALYSIS AND NETWORK RECONSTRUCTION

The developmental processes and functions of an organism are controlled by the genes and the proteins that are derived from these genes. The identification of key genes and the reconstruction of gene networks can provide a model to help us understand the regulatory mechanisms for the initiation and progression of biological processes or functional abnormalities (e.g. diseases) in living organisms. In this dissertation, I have developed statistical methods to identify the genes and transcription factors (TFs) involved in biological processes, constructed their regulatory networks, and also evaluated some existing association methods to find robust methods for coexpression analyses. Two kinds of data sets were used for this work: genotype data and gene expression microarray data. On the basis of these data sets, this dissertation has two major parts, together forming six chapters. The first part deals with developing association methods for rare variants using genotype data (chapter 4 and 5). The second part deals with developing and/or evaluating statistical methods to identify genes and TFs involved in biological processes, and construction of their regulatory networks using gene expression data (chapter 2, 3, and 6). For the first part, I have developed two methods to find the groupwise association of rare variants with given diseases or traits. The first method is based on kernel machine learning and can be applied to both quantitative as well as qualitative traits. Simulation results showed that the proposed method has improved power over the existing weighted sum method (WS) in most settings. The second method uses multiple phenotypes to select a few top significant genes. It then finds the association of each gene with each phenotype while controlling the population stratification by adjusting the data for ancestry using principal components. This method was applied to GAW 17 data and was able to find several disease risk genes. For the second part, I have worked on three problems. First problem involved evaluation of eight gene association methods. A very comprehensive comparison of these methods with further analysis clearly demonstrates the distinct and common performance of these eight gene association methods. For the second problem, an algorithm named the bottom-up graphical Gaussian model was developed to identify the TFs that regulate pathway genes and reconstruct their hierarchical regulatory networks. This algorithm has produced very significant results and it is the first report to produce such hierarchical networks for these pathways. The third problem dealt with developing another algorithm called the top-down graphical Gaussian model that identifies the network governed by a specific TF. The network produced by the algorithm is proven to be of very high accuracy.

Hierarchical Markov Random Fields Applied to Model Soft Tissue Deformations on Graphics Hardware

Many methodologies dealing with prediction or simulation of soft tissue deformations on medical image data require preprocessing of the data in order to produce a different shape representation that complies with standard methodologies, such as mass–spring networks, finite element method s (FEM). On the other hand, methodologies working directly on the image space normally do not take into account mechanical behavior of tissues and tend to lack physics foundations driving soft tissue deformations. This chapter presents a method to simulate soft tissue deformations based on coupled concepts from image analysis and mechanics theory. The proposed methodology is based on a robust stochastic approach that takes into account material properties retrieved directly from the image, concepts from continuum mechanics and FEM. The optimization framework is solved within a hierarchical Markov random field (HMRF) which is implemented on the graphics processor unit (GPU See Graphics processing unit ).

Hierarchical modeling gave plausible estimates of associations between metabolic syndrome and components of antiretroviral therapy

OBJECTIVE: Hierarchical modeling has been proposed as a solution to the multiple exposure problem. We estimate associations between metabolic syndrome and different components of antiretroviral therapy using both conventional and hierarchical models. STUDY DESIGN AND SETTING: We use discrete time survival analysis to estimate the association between metabolic syndrome and cumulative exposure to 16 antiretrovirals from four drug classes. We fit a hierarchical model where the drug class provides a prior model of the association between metabolic syndrome and exposure to each antiretroviral. RESULTS: One thousand two hundred and eighteen patients were followed for a median of 27 months, with 242 cases of metabolic syndrome (20%) at a rate of 7.5 cases per 100 patient years. Metabolic syndrome was more likely to develop in patients exposed to stavudine, but was less likely to develop in those exposed to atazanavir. The estimate for exposure to atazanavir increased from hazard ratio of 0.06 per 6 months' use in the conventional model to 0.37 in the hierarchical model (or from 0.57 to 0.81 when using spline-based covariate adjustment). CONCLUSION: These results are consistent with trials that show the disadvantage of stavudine and advantage of atazanavir relative to other drugs in their respective classes. The hierarchical model gave more plausible results than the equivalent conventional model.

Path Constitution Analysis: A Methodology for Understanding Path Dependence and Path Creation

Although an increasing number of studies of technological, institutional and organizational change refer to the concepts of path dependence and path creation, few attempts have been made to consider these concepts explicitly in their methodological accounts. This paper addresses this gap and contributes to the literature by developing a comprehensive methodology that originates from the concepts of path dependence and path creation – path constitution analysis (PCA) – and allows for the integration of multi-actor constellations on multiple levels of analysis within a process perspective. Based upon a longitudinal case study in the field of semiconductors, we illustrate PCA ‘in action’ as a template for other researchers and critically examine its adequacy. We conclude with implications for further path-oriented inquiries.

Is the relation between early post-session reports and treatment outcome an epiphenomenon of intake distress and early response? A multi-predictor analysis in outpatient psychotherapy

The early phase of psychotherapy has been regarded as a sensitive period in the unfolding of psychotherapy leading to positive outcomes. However, there is disagreement about the degree to which early (especially relationship-related) session experiences predict outcome over and above initial levels of distress and early response to treatment. The goal of the present study was to simultaneously examine outcome at post treatment as a function of (a) intake symptom and interpersonal distress as well as early change in well-being and symptoms, (b) the patient's early session-experiences, (c) the therapist's early session-experiences/interventions, and (d) their interactions. The data of 430 psychotherapy completers treated by 151 therapists were analyzed using hierarchical linear models. Results indicate that early positive intra- and interpersonal session experiences as reported by patients and therapists after the sessions explained 58% of variance of a composite outcome measure, taking intake distress and early response into account. All predictors (other than problem-activating therapists' interventions) contributed to later treatment outcomes if entered as single predictors. However, the multi-predictor analyses indicated that interpersonal distress at intake as well as the early interpersonal session experiences by patients and therapists remained robust predictors of outcome. The findings underscore that early in therapy therapists (and their supervisors) need to understand and monitor multiple interconnected components simultaneously

A 350 ka record of climate change from Lake El'gygytgyn, Far East Russian Arctic: refining the pattern of climate modes by means of cluster analysis

Abstract. Rock magnetic, biochemical and inorganic records of the sediment cores PG1351 and Lz1024 from Lake El’gygytgyn, Chukotka peninsula, Far East Russian Arctic, were subject to a hierarchical agglomerative cluster analysis in order to refine and extend the pattern of climate modes as defined by Melles et al. (2007). Cluster analysis of the data obtained from both cores yielded similar results, differentiating clearly between the four climate modes warm, peak warm, cold and dry, and cold and moist. In addition, two transitional phases were identified, representing the early stages of a cold phase and slightly colder conditions during a warm phase. The statistical approach can thus be used to resolve gradual changes in the sedimentary units as an indicator of available oxygen in the hypolimnion in greater detail. Based upon cluster analyses on core Lz1024, the published succession of climate modes in core PG1351, covering the last 250 ka, was modified and extended back to 350 ka. Comparison to the marine oxygen isotope (�18O) stack LR04 (Lisiecki and Raymo, 2005) and the summer insolation at 67.5� N, with the extended Lake El’gygytgyn parameter records of magnetic susceptibility (�LF), total organic carbon content (TOC) and the chemical index of alteration (CIA; Minyuk et al., 2007), revealed that all stages back to marine isotope stage (MIS) 10 and most of the substages are clearly reflected in the pattern derived from the cluster analysis.

Two decades of measuring environmental attitudes: A comparative analysis of 33 countries

This paper analyzes the development of environmental concern by using the three waves of the environmental modules of the International Social Survey Programme. First, we discuss the measurement of environmental concern and construct a ranking of countries according to the new 2010 results. Second, we analyze the determinants of environmental concern by employing multilevel models that take individual as well as context effects into account. Third, we explore the longitudinal aspect of the data at the macro level in order to uncover the causal relation between countries’ wealth and environmental concern. The results show that environmental concern is closely correlated with the wealth of the nations. However, environmental concern decreased in almost all nations slightly during the last two decades. The decline was lower in countries with improving economic conditions suggesting that economic growth helps to maintain higher levels of environmental concern.

Hierarchical segmentation-assisted multimodal registration for MR brain images

Information theory-based metric such as mutual information (MI) is widely used as similarity measurement for multimodal registration. Nevertheless, this metric may lead to matching ambiguity for non-rigid registration. Moreover, maximization of MI alone does not necessarily produce an optimal solution. In this paper, we propose a segmentation-assisted similarity metric based on point-wise mutual information (PMI). This similarity metric, termed SPMI, enhances the registration accuracy by considering tissue classification probabilities as prior information, which is generated from an expectation maximization (EM) algorithm. Diffeomorphic demons is then adopted as the registration model and is optimized in a hierarchical framework (H-SPMI) based on different levels of anatomical structure as prior knowledge. The proposed method is evaluated using Brainweb synthetic data and clinical fMRI images. Both qualitative and quantitative assessment were performed as well as a sensitivity analysis to the segmentation error. Compared to the pure intensity-based approaches which only maximize mutual information, we show that the proposed algorithm provides significantly better accuracy on both synthetic and clinical data.