Direct estimation and inference of higher-level correlations from lower-level measurements with applications in gene-pathway and proteomics studies.

This paper tackles the challenge of estimating correlations between higher-level biological variables (e.g. proteins and gene pathways) when only lower-level measurements are directly observed (e.g. peptides and individual genes). Existing methods typically aggregate lower-level data into higher-level variables and then estimate correlations based on the aggregated data. However, different data aggregation methods can yield varying correlation estimates as they target different higher-level quantities. Our solution is a latent factor model that directly estimates these higher-level correlations from lower-level data without the need for data aggregation. We further introduce a shrinkage estimator to ensure the positive definiteness and improve the accuracy of the estimated correlation matrix. Furthermore, we establish the asymptotic normality of our estimator, enabling efficient computation of P-values for the identification of significant correlations. The effectiveness of our approach is demonstrated through comprehensive simulations and the analysis of proteomics and gene expression datasets. We develop the R package highcor for implementing our method.

DifferentialRegulation: a Bayesian hierarchical approach to identify differentially regulated genes.

Although transcriptomics data is typically used to analyze mature spliced mRNA, recent attention has focused on jointly investigating spliced and unspliced (or precursor-) mRNA, which can be used to study gene regulation and changes in gene expression production. Nonetheless, most methods for spliced/unspliced inference (such as RNA velocity tools) focus on individual samples, and rarely allow comparisons between groups of samples (e.g. healthy vs. diseased). Furthermore, this kind of inference is challenging, because spliced and unspliced mRNA abundance is characterized by a high degree of quantification uncertainty, due to the prevalence of multi-mapping reads, ie reads compatible with multiple transcripts (or genes), and/or with both their spliced and unspliced versions. Here, we present DifferentialRegulation, a Bayesian hierarchical method to discover changes between experimental conditions with respect to the relative abundance of unspliced mRNA (over the total mRNA). We model the quantification uncertainty via a latent variable approach, where reads are allocated to their gene/transcript of origin, and to the respective splice version. We designed several benchmarks where our approach shows good performance, in terms of sensitivity and error control, vs. state-of-the-art competitors. Importantly, our tool is flexible, and works with both bulk and single-cell RNA-sequencing data. DifferentialRegulation is distributed as a Bioconductor R package.

High response diversity and conspecific density-dependence, not species interactions, drive dynamics of coral reef fish communities.

Species-to-species and species-to-environment interactions are key drivers of community dynamics. Disentangling these drivers in species-rich assemblages is challenging due to the high number of potentially interacting species (the 'curse of dimensionality'). We develop a process-based model that quantifies how intraspecific and interspecific interactions, and species' covarying responses to environmental fluctuations, jointly drive community dynamics. We fit the model to reef fish abundance time series from 41 reefs of Australia's Great Barrier Reef. We found that fluctuating relative abundances are driven by species' heterogenous responses to environmental fluctuations, whereas interspecific interactions are negligible. Species differences in long-term average abundances are driven by interspecific variation in the magnitudes of both conspecific density-dependence and density-independent growth rates. This study introduces a novel approach to overcoming the curse of dimensionality, which reveals highly individualistic dynamics in coral reef fish communities that imply a high level of niche structure.

Quantifying reciprocal relationships between poverty and health: combining a causal loop diagram with longitudinal structural equation modelling.

BACKGROUND: This study takes on the challenge of quantifying a complex causal loop diagram describing how poverty and health affect each other, and does so using longitudinal data from The Netherlands. Furthermore, this paper elaborates on its methodological approach in order to facilitate replication and methodological advancement. METHODS: After adapting a causal loop diagram that was built by stakeholders, a longitudinal structural equation modelling approach was used. A cross-lagged panel model with nine endogenous variables, of which two latent variables, and three time-invariant exogenous variables was constructed. With this model, directional effects are estimated in a Granger-causal manner, using data from 2015 to 2019. Both the direct effects (with a one-year lag) and total effects over multiple (up to eight) years were calculated. Five sensitivity analyses were conducted. Two of these focus on lower-income and lower-wealth individuals. The other three each added one exogenous variable: work status, level of education, and home ownership. RESULTS: The effects of income and financial wealth on health are present, but are relatively weak for the overall population. Sensitivity analyses show that these effects are stronger for those with lower incomes or wealth. Physical capability does seem to have strong positive effects on both income and financial wealth. There are a number of other results as well, as the estimated models are extensive. Many of the estimated effects only become substantial after several years. CONCLUSIONS: Income and financial wealth appear to have limited effects on the health of the overall population of The Netherlands. However, there are indications that these effects may be stronger for individuals who are closer to the poverty threshold. Since the estimated effects of physical capability on income and financial wealth are more substantial, a broad recommendation would be that including physical capability in efforts that are aimed at improving income and financial wealth could be useful and effective. The methodological approach described in this paper could also be applied to other research settings or topics.

Observing the neighborhood multidimensionality and overweight in children of different developmental stages-the emergence of an unusual association.

The present study examines whether the association of the neighborhood environment and overweight in children is moderated by age. This was a cross-sectional study of 832 children aged 3 to 10 years living in the city of Oporto (Portugal). Children were recruited under the scope of the project "Inequalities in Childhood Obesity: The impact of the socioeconomic crisis in Portugal from 2009 to 2015." Overweight was defined according to the International Obesity Task Force criteria. Parents completed a self-administered questionnaire capturing sociodemographic characteristics and their perceptions of their neighborhood environment. Logistic regressions were used to examine the influence of parental perceived neighborhood characteristics (latent variables: attractiveness, traffic safety, crime safety, and walkability) on overweight in children. A stratified analysis by age category was conducted. Overall, 27.8% of the children were overweight, 17.4% were aged 3 to 5 years, and 31.8% were aged 6 to 10 years. Children aged 3 to 5 years were more sensitive to the neighborhood environment than children aged 6 to 10 years. For children aged 3 to 5 years, the risk of overweight was inversely associated with neighborhood crime safety (OR = 1.84; 95% CI 1.07-3.15; p = 0.030).    Conclusion: Our study suggests the existence of a sensitive age period in childhood at which exposure to a hostile neighborhood environment is most determining for weight gain. Until today, it was thought that the impact of the neighborhood environment on younger children would be less important as they are less autonomous. But it may not be true. What is Known: • The neighborhood environment may adversely affect children's weight status. However, the moderating role of child age in the association between neighborhood environment and overweight is uncertain. What is New: • The study highlights that the association between the neighborhood environment and child overweight is attenuated by age. It is stronger for preschoolers than for early school-age children.

Assessing the incorporation of latent variables in the estimation of the value of a statistical life.

For many years, the economic literature has recognized the role of attitudes, beliefs, and perceptions in estimating the value of a statistical life (VSL). However, few applications have attempted to include them. This article incorporates the perceived controllability and concern about traffic and cardiorespiratory risks to estimate VSL using a hybrid choice model (HCM). The HCM allows us to include unobserved heterogeneity and improve behavioral realism explicitly. Using data from a choice experiment conducted in Santiago, Chile, we estimate a VSL of US$3.78 million for traffic risks and US$2.06 million for cardiorespiratory risks. We found that higher controllability decreases the likelihood that the respondents would be willing to pay for risk reductions in both risks. On the other hand, concern about these risks decreases the willingness to pay for traffic risk reductions but increases it for cardiorespiratory risk reductions.

Measurement error models with zero inflation and multiple sources of zeros, with applications to hard zeros.

We consider measurement error models for two variables observed repeatedly and subject to measurement error. One variable is continuous, while the other variable is a mixture of continuous and zero measurements. This second variable has two sources of zeros. The first source is episodic zeros, wherein some of the measurements for an individual may be zero and others positive. The second source is hard zeros, i.e., some individuals will always report zero. An example is the consumption of alcohol from alcoholic beverages: some individuals consume alcoholic beverages episodically, while others never consume alcoholic beverages. However, with a small number of repeat measurements from individuals, it is not possible to determine those who are episodic zeros and those who are hard zeros. We develop a new measurement error model for this problem, and use Bayesian methods to fit it. Simulations and data analyses are used to illustrate our methods. Extensions to parametric models and survival analysis are discussed briefly.

Investigating weight constraint methods for causal-formative indicator modeling.

Causal-formative indicators are often used in social science research. To achieve identification in causal-formative indicator modeling, constraints need to be applied. A conventional method is to constrain the weight of a formative indicator to be 1. The selection of which indicator to have the fixed weight, however, may influence statistical inferences of the structural path coefficients from the causal-formative construct to outcomes. Another conventional method is to use equal weights (e.g., 1) and assumes that all indicators equally contribute to the latent construct, which can be a strong assumption. To address the limitations of the conventional methods, we proposed an alternative constraint method, in which the sum of the weights is constrained to be a constant. We analytically studied the relations and interpretations of structural path coefficients from the constraint methods, and the results showed that the proposed method yields better interpretations of path coefficients. Simulation studies were conducted to compare the performance of the weight constraint methods in causal-formative indicator modeling with one or two outcomes. Results showed that higher biases in the path coefficient estimates were observed from the conventional methods compared to the proposed method. The proposed method had ignorable bias and satisfactory coverage rates in the studied conditions. This study emphasizes the importance of using an appropriate weight constraint method in causal-formative indicator modeling.

Estimation and visualization of heterogeneous treatment effects for multiple outcomes.

We consider two-arm comparison in clinical trials. The objective is to identify a population with characteristics that make the treatment effective. Such a population is called a subgroup. This identification can be made by estimating the treatment effect and identifying the interactions between treatments and covariates. For a single outcome, there are several ways available to identify the subgroups. There are also multiple outcomes, but they are difficult to interpret and cannot be applied to outcomes other than continuous values. In this paper, we thus propose a new method that allows for a straightforward interpretation of subgroups and deals with both continuous and binary outcomes. The proposed method introduces latent variables and adds Lasso sparsity constraints to the estimated loadings to facilitate the interpretation of the relationship between outcomes and covariates. The interpretation of the subgroups is made by visualizing treatment effects and latent variables. Since we are performing sparse estimation, we can interpret the covariates related to the treatment effects and subgroups. Finally, simulation and real data examples demonstrate the effectiveness of the proposed method.

Supervised structural learning of semiparametric regression on high-dimensional correlated covariates with applications to eQTL studies.

Expression quantitative trait loci (eQTL) studies utilize regression models to explain the variance of gene expressions with genetic loci or single nucleotide polymorphisms (SNPs). However, regression models for eQTL are challenged by the presence of high dimensional non-sparse and correlated SNPs with small effects, and nonlinear relationships between responses and SNPs. Principal component analyses are commonly conducted for dimension reduction without considering responses. Because of that, this non-supervised learning method often does not work well when the focus is on discovery of the response-covariate relationship. We propose a new supervised structural dimensional reduction method for semiparametric regression models with high dimensional and correlated covariates; we extract low-dimensional latent features from a vast number of correlated SNPs while accounting for their relationships, possibly nonlinear, with gene expressions. Our model identifies important SNPs associated with gene expressions and estimates the association parameters via a likelihood-based algorithm. A GTEx data application on a cancer related gene is presented with 18 novel eQTLs detected by our method. In addition, extensive simulations show that our method outperforms the other competing methods in bias, efficiency, and computational cost.

Toward Advancing Precision Environmental Health: Developing a Customized Exposure Burden Score to PFAS Mixtures to Enable Equitable Comparisons Across Population Subgroups, Using Mixture Item Response Theory.

Quantifying a person's cumulative exposure burden to per- and polyfluoroalkyl substances (PFAS) mixtures is important for risk assessment, biomonitoring, and reporting of results to participants. However, different people may be exposed to different sets of PFASs due to heterogeneity in the exposure sources and patterns. Applying a single measurement model for the entire population (e.g., by summing concentrations of all PFAS analytes) assumes that each PFAS analyte is equally informative to PFAS exposure burden for all individuals. This assumption may not hold if PFAS exposure sources systematically differ within the population. However, the sociodemographic, dietary, and behavioral characteristics that underlie systematic exposure differences may not be known, or may be due to a combination of these factors. Therefore, we used mixture item response theory, an unsupervised psychometrics and data science method, to develop a customized PFAS exposure burden scoring algorithm. This scoring algorithm ensures that PFAS burden scores can be equitably compared across population subgroups. We applied our methods to PFAS biomonitoring data from the United States National Health and Nutrition Examination Survey (2013-2018). Using mixture item response theory, we found that participants with higher household incomes had higher PFAS burden scores. Asian Americans had significantly higher PFAS burden compared with non-Hispanic Whites and other race/ethnicity groups. However, some disparities were masked when using summed PFAS concentrations as the exposure metric. This work demonstrates that our summary PFAS burden metric, accounting for sources of exposure variation, may be a more fair and informative estimate of PFAS exposure.

Diversifying Emotional Dialogue Generation via Selective Adversarial Training.

Emotional perception and expression are very important for building intelligent conversational systems that are human-like and attractive. Although deep neural approaches have made great progress in the field of conversation generation, there is still a lot of room for research on how to guide systems in generating responses with appropriate emotions. Meanwhile, the problem of systems' tendency to generate high-frequency universal responses remains largely unsolved. To solve this problem, we propose a method to generate diverse emotional responses through selective perturbation. Our model includes a selective word perturbation module and a global emotion control module. The former is used to introduce disturbance factors into the generated responses and enhance their expression diversity. The latter maintains the coherence of the response by limiting the emotional distribution of the response and preventing excessive deviation of emotion and meaning. Experiments are designed on two datasets, and corresponding results show that our model outperforms existing baselines in terms of emotional expression and response diversity.

Inference in Gaussian state-space models with mixed effects for multiple epidemic dynamics.

The estimation from available data of parameters governing epidemics is a major challenge. In addition to usual issues (data often incomplete and noisy), epidemics of the same nature may be observed in several places or over different periods. The resulting possible inter-epidemic variability is rarely explicitly considered. Here, we propose to tackle multiple epidemics through a unique model incorporating a stochastic representation for each epidemic and to jointly estimate its parameters from noisy and partial observations. By building on a previous work for prevalence data, a Gaussian state-space model is extended to a model with mixed effects on the parameters describing simultaneously several epidemics and their observation process. An appropriate inference method is developed, by coupling the SAEM algorithm with Kalman-type filtering. Moreover, we consider here incidence data, which requires to develop a new version of the filtering algorithm. Its performances are investigated on SIR simulated epidemics for prevalence and incidence data. Our method outperforms an inference method separately processing each dataset. An application to SEIR influenza outbreaks in France over several years using incidence data is also carried out. Parameter estimations highlight a non-negligible variability between influenza seasons, both in transmission and case reporting. The main contribution of our study is to rigorously and explicitly account for the inter-epidemic variability between multiple outbreaks, both from the viewpoint of modeling and inference with a parsimonious statistical model.

Repeated Measures ANOVA with Latent Variables to Analyze Interindividual Differences in Contrasts.

Repeated measures analysis of variance (RM-ANOVA) is a broadly used statistical method to analyze data from experimental designs. RM-ANOVA aims at investigating effects of experimental conditions (i.e., factors) and predictors that affect the outcome of interest. It mainly considers contrasts that test standard main and interaction effects, even though more complex contrasts can in principle be used. Analyses, however, only focus on drawing conclusions about average effects and do not take into consideration interindividual differences in these effects. We propose an alternative approach to RM-ANOVA for analyzing repeated measures data, termed latent repeated measures analysis of variance (L-RM-ANOVA). The new approach is based on structural equation modeling and extends the latent growth components approach. L-RM-ANOVA enables the researcher to not only consider mean differences between different experimental conditions (i.e., average effects), but also to investigate interindividual differences in effects. Such interindividual differences are considered with regard to standard main and interactions effects and also with regard to customized contrasts that allow for testing specific hypotheses of interest. Furthermore, L-RM-ANOVA can include a measurement model for latent variables and can be used for the analysis of complex multi-factorial repeated measures designs. We conclude the presentation by demonstrating L-RM-ANOVA using a minimal repeated measures example.

Joint Hidden Markov Model for Longitudinal and Time-to-Event Data with Latent Variables.

This study develops a new joint modeling approach to simultaneously analyze longitudinal and time-to-event data with latent variables. The proposed model consists of three components. The first component is a hidden Markov model for investigating a longitudinal observation process and its underlying transition process as well as their potential risk factors and dynamic heterogeneity. The second component is a factor analysis model for characterizing latent risk factors through multiple observed variables. The third component is a proportional hazards model for examining the effects of observed and latent risk factors on the hazards of interest. A shared random effect is introduced to allow the longitudinal and time-to-event outcomes to be correlated. A Bayesian approach coupled with efficient Markov chain Monte Carlo methods is developed to conduct statistical inference. The performance of the proposed method is evaluated through simulation studies. An application of the proposed model to a general health survey study concerning cognitive impairment and mortality for Chinese elders is presented.

Bayesian nonparametric dynamic hazard rates in evolutionary life tables.

In the study of life tables the random variable of interest is usually assumed discrete since mortality rates are studied for integer ages. In dynamic life tables a time domain is included to account for the evolution effect of the hazard rates in time. In this article we follow a survival analysis approach and use a nonparametric description of the hazard rates. We construct a discrete time stochastic processes that reflects dependence across age as well as in time. This process is used as a bayesian nonparametric prior distribution for the hazard rates for the study of evolutionary life tables. Prior properties of the process are studied and posterior distributions are derived. We present a simulation study, with the inclusion of right censored observations, as well as a real data analysis to show the performance of our model.

Estimating and inferring the maximum degree of stimulus-locked time-varying brain connectivity networks.

Neuroscientists have enjoyed much success in understanding brain functions by constructing brain connectivity networks using data collected under highly controlled experimental settings. However, these experimental settings bear little resemblance to our real-life experience in day-to-day interactions with the surroundings. To address this issue, neuroscientists have been measuring brain activity under natural viewing experiments in which the subjects are given continuous stimuli, such as watching a movie or listening to a story. The main challenge with this approach is that the measured signal consists of both the stimulus-induced signal, as well as intrinsic-neural and nonneuronal signals. By exploiting the experimental design, we propose to estimate stimulus-locked brain networks by treating nonstimulus-induced signals as nuisance parameters. In many neuroscience applications, it is often important to identify brain regions that are connected to many other brain regions during cognitive process. We propose an inferential method to test whether the maximum degree of the estimated network is larger than a prespecific number. We prove that the type I error can be controlled and that the power increases to one asymptotically. Simulation studies are conducted to assess the performance of our method. Finally, we analyze a functional magnetic resonance imaging dataset obtained under the Sherlock Holmes movie stimuli.

Composite likelihood inference for ordinal periodontal data with replicated spatial patterns.

Spatial ordinal data observed separately for multiple subjects are common in biomedical research, yet statistical methodology for such ordinal data analysis is limited. The existing methodology often assumes a single realization of spatial ordinal data without replications, a commonplace in disease mapping studies, and thus are not directly applicable. Motivated by a dataset evaluating periodontal disease (PD) status, we propose a multisubject spatial ordinal model that assumes a geostatistical spatial structure within a regression framework through an elegant latent variable representation. For achieving computational scalability within a classical inferential framework, we develop a maximum composite likelihood method for parameter estimation, and establish the asymptotic properties of the parameter estimates. Another major contribution is the development of model diagnostic measures for our dependent data scenario using generalized surrogate residuals. A simulation study suggests sound finite sample properties of the proposed methods. We also illustrate our proposed methodology via application to the motivating PD dataset. A companion R package clordr is available for easy implementation.

Mixture proportional hazards cure model with latent variables.

A mixture proportional hazards cure model with latent variables is proposed. The proposed model assesses the effects of the observed and latent risk factors on the hazards of uncured subjects and the cure rate through a proportional hazards model and a logistic model, respectively. Factor analysis is employed to measure the latent variables through correlated multiple indicators. Maximum likelihood estimation is performed through a Gaussian quadratic technique that approximates the integration over the latent variables. A piecewise constant function is used for the unspecified baseline hazard of uncured subjects. The proposed method can be conveniently implemented by using SAS Proc NLMIXED. Simulation studies are conducted to evaluate the performance of the proposed approach. An application to a study concerning the risk factors of chronic kidney disease for type 2 diabetic patients is provided.

The effect of species, size, and fire intensity on tree mortality within a catastrophic bushfire complex.

Infrequent, high-intensity disturbances can have profound impacts on forested landscapes, changing forest structure and altering relative species abundance. However, due to their rarity and the logistical challenges of directly observing such extreme events, both the spatial variability of disturbance intensity and the species-specific responses to this variability are poorly understood. We used observed patterns of mortality across a fire severity gradient following the 2009 Black Saturday fires in southeastern Australia to simultaneously estimate (1) species- and size-specific susceptibility to fire-induced mortality and (2) fire intensity. We found broad variation in patterns of fire susceptibility among the 10 tree species (five eucalypts and five non-eucalypts) sufficiently abundant for analysis. Among the eucalypts, Eucalyptus obliqua was the most resistant to fire-induced mortality, with trees of ~25 cm DBH having a 50% probability of surviving even the most intense fires. In contrast, E. regnans had 100% mortality across all size classes when subjected to high-intensity fire. Basal resprouting occurred in six of the study species and, when accounted for, fundamentally changed the mortality profile of these species, highlighting the importance of resprouting as an adaptation to fire in these landscapes. In particular, the two iconic cool temperate rainforest species (Nothofagus cunninghami and Atherosperma moschatum) were strong resprouters (~45% of individuals were able to resprout after being top-killed by fire). We also found evidence for compositional shifts in regeneration above threshold values of fire intensity in cool temperate rainforest and mixed forest sites, both of which have important conservation values within these landscapes. The observed patterns of species- and size-specific susceptibility to fire-induced mortality may be used to anticipate changes in forest structure and composition in the future. In addition, they may also help guide forest management strategies that reduce the length of time individual trees are exposed to potentially lethal fires, thereby increasing the resilience of these forests to future fires.