Bayesian Hidden Markov Modeling of Array CGH Data

Genomic alterations have been linked to the development and progression of cancer. The technique of Comparative Genomic Hybridization (CGH) yields data consisting of fluorescence intensity ratios of test and reference DNA samples. The intensity ratios provide information about the number of copies in DNA. Practical issues such as the contamination of tumor cells in tissue specimens and normalization errors necessitate the use of statistics for learning about the genomic alterations from array-CGH data. As increasing amounts of array CGH data become available, there is a growing need for automated algorithms for characterizing genomic profiles. Specifically, there is a need for algorithms that can identify gains and losses in the number of copies based on statistical considerations, rather than merely detect trends in the data. We adopt a Bayesian approach, relying on the hidden Markov model to account for the inherent dependence in the intensity ratios. Posterior inferences are made about gains and losses in copy number. Localized amplifications (associated with oncogene mutations) and deletions (associated with mutations of tumor suppressors) are identified using posterior probabilities. Global trends such as extended regions of altered copy number are detected. Since the posterior distribution is analytically intractable, we implement a Metropolis-within-Gibbs algorithm for efficient simulation-based inference. Publicly available data on pancreatic adenocarcinoma, glioblastoma multiforme and breast cancer are analyzed, and comparisons are made with some widely-used algorithms to illustrate the reliability and success of the technique.

A unified approach to modeling multivariate binary data using copulas over partitions

Many seemingly disparate approaches for marginal modeling have been developed in recent years. We demonstrate that many current approaches for marginal modeling of correlated binary outcomes produce likelihoods that are equivalent to the proposed copula-based models herein. These general copula models of underlying latent threshold random variables yield likelihood based models for marginal fixed effects estimation and interpretation in the analysis of correlated binary data. Moreover, we propose a nomenclature and set of model relationships that substantially elucidates the complex area of marginalized models for binary data. A diverse collection of didactic mathematical and numerical examples are given to illustrate concepts.

Modeling multilevel sleep transitional data via Poisson log-linear multilevel models

This paper proposes Poisson log-linear multilevel models to investigate population variability in sleep state transition rates. We specifically propose a Bayesian Poisson regression model that is more flexible, scalable to larger studies, and easily fit than other attempts in the literature. We further use hierarchical random effects to account for pairings of individuals and repeated measures within those individuals, as comparing diseased to non-diseased subjects while minimizing bias is of epidemiologic importance. We estimate essentially non-parametric piecewise constant hazards and smooth them, and allow for time varying covariates and segment of the night comparisons. The Bayesian Poisson regression is justified through a re-derivation of a classical algebraic likelihood equivalence of Poisson regression with a log(time) offset and survival regression assuming piecewise constant hazards. This relationship allows us to synthesize two methods currently used to analyze sleep transition phenomena: stratified multi-state proportional hazards models and log-linear models with GEE for transition counts. An example data set from the Sleep Heart Health Study is analyzed.

On Time Series Analysis of Public Health and Biomedical Data

A time series is a sequence of observations made over time. Examples in public health include daily ozone concentrations, weekly admissions to an emergency department or annual expenditures on health care in the United States. Time series models are used to describe the dependence of the response at each time on predictor variables including covariates and possibly previous values in the series. Time series methods are necessary to account for the correlation among repeated responses over time. This paper gives an overview of time series ideas and methods used in public health research.

Anger and posttraumatic stress disorder symptoms in crime victims: A longitudinal analysis

Among trauma-exposed individuals, severity of posttraumatic stress disorder (PTSD) symptoms is strongly correlated with anger. The authors used 2 longitudinal data sets with 282 and 218 crime victims, respectively, to investigate the temporal sequence of anger and PTSD symptoms following the assault. Cross-lagged regression analyses indicated that PTSD symptoms predicted subsequent level of anger, but that anger did not predict subsequent PTSD symptoms. Testing alternative models (common factor model, unmeasured 3rd variable model) that might account for spuriousness of the relation strengthened confidence in the results of the cross-lagged analyses. Further analyses suggested that rumination mediates the effect of PTSD symptoms on anger.

Decreasing risk of hepatitis a infection in león, nicaragua: evidence from cross-sectional and longitudinal seroepidemiology studies

BACKGROUND AND OBJECTIVES Nicaragua is highly endemic for hepatitis A. We aimed to provide an estimate of the change in the age-specific risk of hepatitis A virus (HAV) infection based on serological data from cross-sectional and longitudinal samples collected in León, Nicaragua, in 1995/96 (n = 979) and 2003 (n = 494). METHODS The observed age-specific prevalence of anti-HAV antibodies was correlated to the age-specific risk of infection by calculating the probability of freedom from infection at a specific age. RESULTS The proportion of seropositive children aged 1.5 to 6 years was 42% in 2003 compared to 67% in 1995/96. Estimated annual risk of infection for a 3-year old child was 30% (95% CI: 27.0%, 33.1%) in 1995 and 15.5% (95% CI: 12.4%, 19.0%) in 2003. There was good agreement between estimates based on cross-sectional and longitudinal data. The age-specific geometric mean of the quantified anti-HAV antibody levels assessed in 2003 was highest at age 4 and decreased steadily up to age 40. CONCLUSIONS The substantially lower risk of HAV infection in 2003 than in 1995 for young children indicates a beginning transition from high to intermediate endemicity in León, Nicaragua. Consecutive age-stratified serosurveys are useful to assess changes in risk of infection following public health interventions. The decreasing age-specific GMC of anti-HAV antibodies during adulthood in a country with endemic HAV indirectly suggests that ongoing HAV exposure in the community has marginal boosting effect on antibody levels once protective immunity has been established by natural infection.

Is low self-esteem a risk factor for depression? Findings from a longitudinal study of Mexican-origin youth

We examined the relation between low self-esteem and depression using longitudinal data from a sample of 674 Mexican-origin early adolescents who were assessed at age 10 and 12 years. Results supported the vulnerability model, which states that low self-esteem is a prospective risk factor for depression. Moreover, results suggested that the vulnerability effect of low self-esteem is driven, for the most part, by general evaluations of worth (i.e., global self-esteem), rather than by domain-specific evaluations of academic competence, physical appearance, and competence in peer relationships. The only domain-specific self-evaluation that showed a prospective effect on depression was honesty-trustworthiness. The vulnerability effect of low self-esteem held for male and female adolescents, for adolescents born in the United States versus Mexico, and across different levels of pubertal status. Finally, the vulnerability effect held when we controlled for several theoretically relevant 3rd variables (i.e., social support, maternal depression, stressful events, and relational victimization) and for interactive effects between self-esteem and the 3rd variables. The present study contributes to an emerging understanding of the link between self-esteem and depression and provides much needed data on the antecedents of depression in ethnic minority populations

Does low self-esteem predict depression and anxiety? A meta-analysis of longitudinal studies

Low self-esteem and depression are strongly related, but there is not yet consistent evidence on the nature of the relation. Whereas the vulnerability model states that low self-esteem contributes to depression, the scar model states that depression erodes self-esteem. Furthermore, it is unknown whether the models are specific for depression or whether they are also valid for anxiety. We evaluated the vulnerability and scar models of low self-esteem and depression, and low self-esteem and anxiety, by meta-analyzing the available longitudinal data (covering 77 studies on depression and 18 studies on anxiety). The mean age of the samples ranged from childhood to old age. In the analyses, we used a random-effects model and examined prospective effects between the variables, controlling for prior levels of the predicted variables. For depression, the findings supported the vulnerability model: The effect of self-esteem on depression (β = -.16) was significantly stronger than the effect of depression on self-esteem (β = -.08). In contrast, the effects between low self-esteem and anxiety were relatively balanced: Self-esteem predicted anxiety with β = -.10, and anxiety predicted self-esteem with β = -.08. Moderator analyses were conducted for the effect of low self-esteem on depression; these suggested that the effect is not significantly influenced by gender, age, measures of self-esteem and depression, or time lag between assessments. If future research supports the hypothesized causality of the vulnerability effect of low self-esteem on depression, interventions aimed at increasing self-esteem might be useful in reducing the risk of depression.

Multilevel analysis for a longitudinal clinical trial

This study applies the multilevel analysis technique to longitudinal data of a large clinical trial. The technique accounts for the correlation at different levels when modeling repeated blood pressure measurements taken throughout the trial. This modeling allows for closer inspection of the remaining correlation and non-homogeneity of variance in the data. Three methods of modeling the correlation were compared. ^

The effect of genotype-by-environment interaction on longitudinal triglyceride profiles in the Bogalusa Heart Study

Triglyceride levels are a component of plasma lipids that are thought to be an important risk factor for coronary heart disease and are influenced by genetic and environmental factors, such as single nucleotide polymorphisms (SNPs), alcohol intake, and smoking. This study used longitudinal data from the Bogalusa Heart Study, a biracial community-based survey of cardiovascular disease risk factors. A sample of 1191 individuals, 4 to 38 years of age, was measured multiple times from 1973 to 2000. The study sample consisted of 730 white and 461 African American participants. Individual growth models were developed in order to assess gene-environment interactions affecting plasma triglycerides over time. After testing for inclusion of significant covariates and interactions, final models, each accounting for the effects of a different SNP, were assessed for fit and normality. After adjustment for all other covariates and interactions, LIPC -514C/T was found to interact with age3, age2, and age and a non-significant interaction of CETP -971G/A genotype with smoking status was found (p = 0.0812). Ever-smokers had higher triglyceride levels than never smokers, but persons heterozygous at this locus, about half of both races, had higher triglyceride levels after smoking cessation compared to current smokers. Since tobacco products increase free fatty acids circulating in the bloodstream, smoking cessation programs have the potential to ultimately reduce triglyceride levels for many persons. However, due to the effect of smoking cessation on the triglyceride levels of CETP -971G/A heterozygotes, the need for smoking prevention programs is also demonstrated. Both smoking cessation and prevention programs would have a great public health impact on minimizing triglyceride levels and ultimately reducing heart disease. ^

Quality Data is Key to Improving Education

The Data Quality Campaign (DQC) has been focused since 2005 on advocating for states to build robust state longitudinal data systems (SLDS). While states have made great progress in their data infrastructure, and should continue to emphasize this work, t data systems alone will not improve outcomes. It is time for both DQC and states to focus on building capacity to use the information that these systems are producing at every level – from classrooms to state houses. To impact system performance and student achievement, the ingrained culture must be replaced with one that focuses on data use for continuous improvement. The effective use of data to inform decisions, provide transparency, improve the measurement of outcomes, and fuel continuous improvement will not come to fruition unless there is a system wide focus on building capacity around the collection, analysis, dissemination, and use of this data, including through research.

Multi-level zero-inflated Poisson regression modelling of correlated count data with excess zeros

Count data with excess zeros relative to a Poisson distribution are common in many biomedical applications. A popular approach to the analysis of such data is to use a zero-inflated Poisson (ZIP) regression model. Often, because of the hierarchical Study design or the data collection procedure, zero-inflation and lack of independence may occur simultaneously, which tender the standard ZIP model inadequate. To account for the preponderance of zero counts and the inherent correlation of observations, a class of multi-level ZIP regression model with random effects is presented. Model fitting is facilitated using an expectation-maximization algorithm, whereas variance components are estimated via residual maximum likelihood estimating equations. A score test for zero-inflation is also presented. The multi-level ZIP model is then generalized to cope with a more complex correlation structure. Application to the analysis of correlated count data from a longitudinal infant feeding study illustrates the usefulness of the approach.