Stochastic search variable selection based on two mixture components and continuous‐scale weighting

Stochastic search variable selection (SSVS) is a Bayesian variable selection method that employs covariate‐specific discrete indicator variables to select which covariates (e.g., molecular markers) are included in or excluded from the model. We present a new variant of SSVS where, instead of discrete indicator variables, we use continuous‐scale weighting variables (which take also values between zero and one) to select covariates into the model. The improved model performance is shown and compared to standard SSVS using simulated and real quantitative trait locus mapping datasets. The decision making to decide phenotype‐genotype associations in our SSVS variant is based on median of posterior distribution or using Bayes factors. We also show here that by using continuous‐scale weighting variables it is possible to improve mixing properties of Markov chain Monte Carlo sampling substantially compared to standard SSVS. Also, the separation of association signals and nonsignals (control of noise level) seems to be more efficient compared to the standard SSVS. Thus, the novel method provides efficient new framework for SSVS analysis that additionally provides whole posterior distribution for pseudo‐indicators which means more information and may help in decision making.     

Variable selection – A review and recommendations for the practicing statistician

Statistical models support medical research by facilitating individualized outcome prognostication conditional on independent variables or by estimating effects of risk factors adjusted for covariates. Theory of statistical models is well‐established if the set of independent variables to consider is fixed and small. Hence, we can assume that effect estimates are unbiased and the usual methods for confidence interval estimation are valid. In routine work, however, it is not known a priori which covariates should be included in a model, and often we are confronted with the number of candidate variables in the range 10–30. This number is often too large to be considered in a statistical model. We provide an overview of various available variable selection methods that are based on significance or information criteria, penalized likelihood, the change‐in‐estimate criterion, background knowledge, or combinations thereof. These methods were usually developed in the context of a linear regression model and then transferred to more generalized linear models or models for censored survival data. Variable selection, in particular if used in explanatory modeling where effect estimates are of central interest, can compromise stability of a final model, unbiasedness of regression coefficients, and validity of p‐values or confidence intervals. Therefore, we give pragmatic recommendations for the practicing statistician on application of variable selection methods in general (low‐dimensional) modeling problems and on performing stability investigations and inference. We also propose some quantities based on resampling the entire variable selection process to be routinely reported by software packages offering automated variable selection algorithms.     

Estimating Subject‐Specific Dependent Competing Risk Profile with Censored Event Time Observations

Summary In a longitudinal study, suppose that the primary endpoint is the time to a specific event. This response variable, however, may be censored by an independent censoring variable or by the occurrence of one of several dependent competing events. For each study subject, a set of baseline covariates is collected. The question is how to construct a reliable prediction rule for the future subject's profile of all competing risks of interest at a specific time point for risk‐benefit decision making. In this article, we propose a two‐stage procedure to make inferences about such subject‐specific profiles. For the first step, we use a parametric model to obtain a univariate risk index score system. We then estimate consistently the average competing risks for subjects who have the same parametric index score via a nonparametric function estimation procedure. We illustrate this new proposal with the data from a randomized clinical trial for evaluating the efficacy of a treatment for prostate cancer. The primary endpoint for this study was the time to prostate cancer death, but had two types of dependent competing events, one from cardiovascular death and the other from death of other causes.     

Pairwise Variable Selection for High‐Dimensional Model‐Based Clustering

Summary Variable selection for clustering is an important and challenging problem in high‐dimensional data analysis. Existing variable selection methods for model‐based clustering select informative variables in a “one‐in‐all‐out” manner; that is, a variable is selected if at least one pair of clusters is separable by this variable and removed if it cannot separate any of the clusters. In many applications, however, it is of interest to further establish exactly which clusters are separable by each informative variable. To address this question, we propose a pairwise variable selection method for high‐dimensional model‐based clustering. The method is based on a new pairwise penalty. Results on simulated and real data show that the new method performs better than alternative approaches that use ?₁ and ?_∞ penalties and offers better interpretation.     

A comparison of regression methods for model selection in individual‐based landscape genetic analysis

Anthropogenic migration barriers fragment many populations and limit the ability of species to respond to climate‐induced biome shifts. Conservation actions designed to conserve habitat connectivity and mitigate barriers are needed to unite fragmented populations into larger, more viable metapopulations, and to allow species to track their climate envelope over time. Landscape genetic analysis provides an empirical means to infer landscape factors influencing gene flow and thereby inform such conservation actions. However, there are currently many methods available for model selection in landscape genetics, and considerable uncertainty as to which provide the greatest accuracy in identifying the true landscape model influencing gene flow among competing alternative hypotheses. In this study, we used population genetic simulations to evaluate the performance of seven regression‐based model selection methods on a broad array of landscapes that varied by the number and type of variables contributing to resistance, the magnitude and cohesion of resistance, as well as the functional relationship between variables and resistance. We also assessed the effect of transformations designed to linearize the relationship between genetic and landscape distances. We found that linear mixed effects models had the highest accuracy in every way we evaluated model performance; however, other methods also performed well in many circumstances, particularly when landscape resistance was high and the correlation among competing hypotheses was limited. Our results provide guidance for which regression‐based model selection methods provide the most accurate inferences in landscape genetic analysis and thereby best inform connectivity conservation actions.     

Imputation and variable selection in linear regression models with missing covariates

Across multiply imputed data sets, variable selection methods such as stepwise regression and other criterion-based strategies that include or exclude particular variables typically result in models with different selected predictors, thus presenting a problem for combining the results from separate complete-data analyses. Here, drawing on a Bayesian framework, we propose two alternative strategies to address the problem of choosing among linear regression models when there are missing covariates. One approach, which we call "impute, then select" (ITS) involves initially performing multiple imputation and then applying Bayesian variable selection to the multiply imputed data sets. A second strategy is to conduct Bayesian variable selection and missing data imputation simultaneously within one Gibbs sampling process, which we call "simultaneously impute and select" (SIAS). The methods are implemented and evaluated using the Bayesian procedure known as stochastic search variable selection for multivariate normal data sets, but both strategies offer general frameworks within which different Bayesian variable selection algorithms could be used for other types of data sets. A study of mental health services utilization among children in foster care programs is used to illustrate the techniques. Simulation studies show that both ITS and SIAS outperform complete-case analysis with stepwise variable selection and that SIAS slightly outperforms ITS.     

WFABC: a Wright–Fisher ABC‐based approach for inferring effective population sizes and selection coefficients from time‐sampled data

With novel developments in sequencing technologies, time‐sampled data are becoming more available and accessible. Naturally, there have been efforts in parallel to infer population genetic parameters from these data sets. Here, we compare and analyse four recent approaches based on the Wright–Fisher model for inferring selection coefficients (s) given effective population size (N_e), with simulated temporal data sets. Furthermore, we demonstrate the advantage of a recently proposed approximate Bayesian computation (ABC)‐based method that is able to correctly infer genomewide average N_e from time‐serial data, which is then set as a prior for inferring per‐site selection coefficients accurately and precisely. We implement this ABC method in a new software and apply it to a classical time‐serial data set of the medionigra genotype in the moth Panaxia dominula. We show that a recessive lethal model is the best explanation for the observed variation in allele frequency by implementing an estimator of the dominance ratio (h).     

Development of a risk scoring system for patients with papillary thyroid cancer

As the importance of personalized therapeutics in aggressive papillary thyroid cancer (PTC) increases, accurate risk stratification is required. To develop a novel prognostic scoring system for patients with PTC (n = 455), we used mRNA expression and clinical data from The Cancer Genome Atlas. We performed variable selection using Network‐Regularized high‐dimensional Cox‐regression with gene network from pathway databases. The risk score was calculated using a linear combination of regression coefficients and mRNA expressions. The risk score and clinical variables were assessed by several survival analyses. The risk score showed high discriminatory power for the prediction of event‐free survival as well as the presence of metastasis. In multivariate analysis, the risk score and presence of metastasis were significant risk factors among the clinical variables that were examined together. In the current study, we developed a risk scoring system that will help to identify suitable therapeutic options for PTC.     

Improving inferences from short‐term ecological studies with Bayesian hierarchical modeling: white‐headed woodpeckers in managed forests

Pilot studies are often used to design short‐term research projects and long‐term ecological monitoring programs, but data are sometimes discarded when they do not match the eventual survey design. Bayesian hierarchical modeling provides a convenient framework for integrating multiple data sources while explicitly separating sample variation into observation and ecological state processes. Such an approach can better estimate state uncertainty and improve inferences from short‐term studies in dynamic systems. We used a dynamic multistate occupancy model to estimate the probabilities of occurrence and nesting for white‐headed woodpeckers Picoides albolarvatus in recent harvest units within managed forests of northern California, USA. Our objectives were to examine how occupancy states and state transitions were related to forest management practices, and how the probabilities changed over time. Using Gibbs variable selection, we made inferences using multiple model structures and generated model‐averaged estimates. Probabilities of white‐headed woodpecker occurrence and nesting were high in 2009 and 2010, and the probability that nesting persisted at a site was positively related to the snag density in harvest units. Prior‐year nesting resulted in higher probabilities of subsequent occurrence and nesting. We demonstrate the benefit of forest management practices that increase the density of retained snags in harvest units for providing white‐headed woodpecker nesting habitat. While including an additional year of data from our pilot study did not drastically alter management recommendations, it changed the interpretation of the mechanism behind the observed dynamics. Bayesian hierarchical modeling has the potential to maximize the utility of studies based on small sample sizes while fully accounting for measurement error and both estimation and model uncertainty, thereby improving the ability of observational data to inform conservation and management strategies.     

Competing Risks and Time‐Dependent Covariates

Time‐dependent covariates are frequently encountered in regression analysis for event history data and competing risks. They are often essential predictors, which cannot be substituted by time‐fixed covariates. This study briefly recalls the different types of time‐dependent covariates, as classified by Kalbfleisch and Prentice [The Statistical Analysis of Failure Time Data, Wiley, New York, 2002] with the intent of clarifying their role and emphasizing the limitations in standard survival models and in the competing risks setting. If random (internal) time‐dependent covariates are to be included in the modeling process, then it is still possible to estimate cause‐specific hazards but prediction of the cumulative incidences and survival probabilities based on these is no longer feasible. This article aims at providing some possible strategies for dealing with these prediction problems. In a multi‐state framework, a first approach uses internal covariates to define additional (intermediate) transient states in the competing risks model. Another approach is to apply the landmark analysis as described by van Houwelingen [Scandinavian Journal of Statistics 2007, 34 , 70–85] in order to study cumulative incidences at different subintervals of the entire study period. The final strategy is to extend the competing risks model by considering all the possible combinations between internal covariate levels and cause‐specific events as final states. In all of those proposals, it is possible to estimate the changes/differences of the cumulative risks associated with simple internal covariates. An illustrative example based on bone marrow transplant data is presented in order to compare the different methods.     

Bayesian joint models with INLA exploring marine mobile predator–prey and competitor species habitat overlap

Understanding spatial physical habitat selection driven by competition and/or predator–prey interactions of mobile marine species is a fundamental goal of spatial ecology. However, spatial counts or density data for highly mobile animals often (1) include excess zeros, (2) have spatial correlation, and (3) have highly nonlinear relationships with physical habitat variables, which results in the need for complex joint spatial models. In this paper, we test the use of Bayesian hierarchical hurdle and zero‐inflated joint models with integrated nested Laplace approximation (INLA), to fit complex joint models to spatial patterns of eight mobile marine species (grey seal, harbor seal, harbor porpoise, common guillemot, black‐legged kittiwake, northern gannet, herring, and sandeels). For each joint model, we specified nonlinear smoothed effect of physical habitat covariates and selected either competing species or predator–prey interactions. Out of a range of six ecologically important physical and biologic variables that are predicted to change with climate change and large‐scale energy extraction, we identified the most important habitat variables for each species and present the relationships between these bio/physical variables and species distributions. In particular, we found that net primary production played a significant role in determining habitat preferences of all the selected mobile marine species. We have shown that the INLA method is well‐suited for modeling spatially correlated data with excessive zeros and is an efficient approach to fit complex joint spatial models with nonlinear effects of covariates. Our approach has demonstrated its ability to define joint habitat selection for both competing and prey–predator species that can be relevant to numerous issues in the management and conservation of mobile marine species.     

Simulation‐selection‐extrapolation: Estimation in high‐dimensional errors‐in‐variables models

Errors‐in‐variables models in high‐dimensional settings pose two challenges in application. First, the number of observed covariates is larger than the sample size, while only a small number of covariates are true predictors under an assumption of model sparsity. Second, the presence of measurement error can result in severely biased parameter estimates, and also affects the ability of penalized methods such as the lasso to recover the true sparsity pattern. A new estimation procedure called SIMulation‐SELection‐EXtrapolation (SIMSELEX) is proposed. This procedure makes double use of lasso methodology. First, the lasso is used to estimate sparse solutions in the simulation step, after which a group lasso is implemented to do variable selection. The SIMSELEX estimator is shown to perform well in variable selection, and has significantly lower estimation error than naive estimators that ignore measurement error. SIMSELEX can be applied in a variety of errors‐in‐variables settings, including linear models, generalized linear models, and Cox survival models. It is furthermore shown in the Supporting Information how SIMSELEX can be applied to spline‐based regression models. A simulation study is conducted to compare the SIMSELEX estimators to existing methods in the linear and logistic model settings, and to evaluate performance compared to naive methods in the Cox and spline models. Finally, the method is used to analyze a microarray dataset that contains gene expression measurements of favorable histology Wilms tumors.     

Skew‐normal/independent linear mixed models for censored responses with applications to HIV viral loads

Often in biomedical studies, the routine use of linear mixed‐effects models (based on Gaussian assumptions) can be questionable when the longitudinal responses are skewed in nature. Skew‐normal/elliptical models are widely used in those situations. Often, those skewed responses might also be subjected to some upper and lower quantification limits (QLs; viz., longitudinal viral‐load measures in HIV studies), beyond which they are not measurable. In this paper, we develop a Bayesian analysis of censored linear mixed models replacing the Gaussian assumptions with skew‐normal/independent (SNI) distributions. The SNI is an attractive class of asymmetric heavy‐tailed distributions that includes the skew‐normal, skew‐t, skew‐slash, and skew‐contaminated normal distributions as special cases. The proposed model provides flexibility in capturing the effects of skewness and heavy tail for responses that are either left‐ or right‐censored. For our analysis, we adopt a Bayesian framework and develop a Markov chain Monte Carlo algorithm to carry out the posterior analyses. The marginal likelihood is tractable, and utilized to compute not only some Bayesian model selection measures but also case‐deletion influence diagnostics based on the Kullback–Leibler divergence. The newly developed procedures are illustrated with a simulation study as well as an HIV case study involving analysis of longitudinal viral loads.     

Subset selection for linear mixed models

Linear mixed models (LMMs) are instrumental for regression analysis with structured dependence, such as grouped, clustered, or multilevel data. However, selection among the covariates—while accounting for this structured dependence—remains a challenge. We introduce a Bayesian decision analysis for subset selection with LMMs. Using a Mahalanobis loss function that incorporates the structured dependence, we derive optimal linear coefficients for (i) any given subset of variables and (ii) all subsets of variables that satisfy a cardinality constraint. Crucially, these estimates inherit shrinkage or regularization and uncertainty quantification from the underlying Bayesian model, and apply for any well-specified Bayesian LMM. More broadly, our decision analysis strategy deemphasizes the role of a single “best” subset, which is often unstable and limited in its information content, and instead favors a collection of near-optimal subsets. This collection is summarized by key member subsets and variable-specific importance metrics. Customized subset search and out-of-sample approximation algorithms are provided for more scalable computing. These tools are applied to simulated data and a longitudinal physical activity dataset, and demonstrate excellent prediction, estimation, and selection ability.     

Variable selection in Bayesian generalized linear‐mixed models: An illustration using candidate gene case‐control association studies

The problem of variable selection in the generalized linear‐mixed models (GLMMs) is pervasive in statistical practice. For the purpose of variable selection, many methodologies for determining the best subset of explanatory variables currently exist according to the model complexity and differences between applications. In this paper, we develop a “higher posterior probability model with bootstrap” (HPMB) approach to select explanatory variables without fitting all possible GLMMs involving a small or moderate number of explanatory variables. Furthermore, to save computational load, we propose an efficient approximation approach with Laplace's method and Taylor's expansion to approximate intractable integrals in GLMMs. Simulation studies and an application of HapMap data provide evidence that this selection approach is computationally feasible and reliable for exploring true candidate genes and gene–gene associations, after adjusting for complex structures among clusters.     

Evolution of risk‐taking during conspicuous mating displays

In many species, males possess conspicuous characteristics to attract females. These traits often attract predators as well, and males thus may have to balance the conspicuousness of their signals in relation to the prevailing predation risk. Here we develop a theoretical model of optimal signaling and risk‐taking behavior for males differing in the attractiveness of their signals. All else being equal, more attractive males should behave more cautiously. Yet this prediction may drastically change if males differ in any additional characteristic, especially if basal mortality rate or signaling costs are higher or if the vulnerability to predators is lower for attractive males. A key insight from our model is that male competition will create a positive feedback so that selection on male risk‐taking strategies is acting in opposite directions. If selection acts on one male type to behave more cautiously, this will strengthen selection on males of the other types to take higher risks and vice versa. Our results further demonstrate that the asset‐protection principle, which states that individuals with higher future expectations should behave more cautiously, may often be violated. We also offer an alternative to the handicap principle explaining the often found positive association between male ornamentation and viability: attractive males may simply behave more cautiously.     

LNRLMI: Linear neighbour representation for predicting lncRNA‐miRNA interactions

LncRNA and miRNA are key molecules in mechanism of competing endogenous RNAs(ceRNA), and their interactions have been discovered with important roles in gene regulation. As supplementary to the identification of lncRNA‐miRNA interactions from CLIP‐seq experiments, in silico prediction can select the most potential candidates for experimental validation. Although developing computational tool for predicting lncRNA‐miRNA interaction is of great importance for deciphering the ceRNA mechanism, little effort has been made towards this direction. In this paper, we propose an approach based on linear neighbour representation to predict lncRNA‐miRNA interactions (LNRLMI). Specifically, we first constructed a bipartite network by combining the known interaction network and similarities based on expression profiles of lncRNAs and miRNAs. Based on such a data integration, linear neighbour representation method was introduced to construct a prediction model. To evaluate the prediction performance of the proposed model, k‐fold cross validations were implemented. As a result, LNRLMI yielded the average AUCs of 0.8475 ± 0.0032, 0.8960 ± 0.0015 and 0.9069 ± 0.0014 on 2‐fold, 5‐fold and 10‐fold cross validation, respectively. A series of comparison experiments with other methods were also conducted, and the results showed that our method was feasible and effective to predict lncRNA‐miRNA interactions via a combination of different types of useful side information. It is anticipated that LNRLMI could be a useful tool for predicting non‐coding RNA regulation network that lncRNA and miRNA are involved in.     

Developing approaches for linear mixed modeling in landscape genetics through landscape‐directed dispersal simulations

Dispersal can impact population dynamics and geographic variation, and thus, genetic approaches that can establish which landscape factors influence population connectivity have ecological and evolutionary importance. Mixed models that account for the error structure of pairwise datasets are increasingly used to compare models relating genetic differentiation to pairwise measures of landscape resistance. A model selection framework based on information criteria metrics or explained variance may help disentangle the ecological and landscape factors influencing genetic structure, yet there are currently no consensus for the best protocols. Here, we develop landscape‐directed simulations and test a series of replicates that emulate independent empirical datasets of two species with different life history characteristics (greater sage‐grouse; eastern foxsnake). We determined that in our simulated scenarios, AIC and BIC were the best model selection indices and that marginal R² values were biased toward more complex models. The model coefficients for landscape variables generally reflected the underlying dispersal model with confidence intervals that did not overlap with zero across the entire model set. When we controlled for geographic distance, variables not in the underlying dispersal models (i.e., nontrue) typically overlapped zero. Our study helps establish methods for using linear mixed models to identify the features underlying patterns of dispersal across a variety of landscapes.     

Bayesian Variable Selection for Multivariate Spatially Varying Coefficient Regression

Summary Physical activity has many well‐documented health benefits for cardiovascular fitness and weight control. For pregnant women, the American College of Obstetricians and Gynecologists currently recommends 30 minutes of moderate exercise on most, if not all, days; however, very few pregnant women achieve this level of activity. Traditionally, studies have focused on examining individual or interpersonal factors to identify predictors of physical activity. There is a renewed interest in whether characteristics of the physical environment in which we live and work may also influence physical activity levels. We consider one of the first studies of pregnant women that examines the impact of characteristics of the built environment on physical activity levels. Using a socioecologic framework, we study the associations between physical activity and several factors including personal characteristics, meteorological/air quality variables, and neighborhood characteristics for pregnant women in four counties of North Carolina. We simultaneously analyze six types of physical activity and investigate cross‐dependencies between these activity types. Exploratory analysis suggests that the associations are different in different regions. Therefore, we use a multivariate regression model with spatially varying regression coefficients. This model includes a regression parameter for each covariate at each spatial location. For our data with many predictors, some form of dimension reduction is clearly needed. We introduce a Bayesian variable selection procedure to identify subsets of important variables. Our stochastic search algorithm determines the probabilities that each covariate's effect is null, non‐null but constant across space, and spatially varying. We found that individual‐level covariates had a greater influence on women's activity levels than neighborhood environmental characteristics, and some individual‐level covariates had spatially varying associations with the activity levels of pregnant women.     

An exploration of fixed and random effects selection for longitudinal binary outcomes in the presence of nonignorable dropout

We explore a Bayesian approach to selection of variables that represent fixed and random effects in modeling of longitudinal binary outcomes with missing data caused by dropouts. We show via analytic results for a simple example that nonignorable missing data lead to biased parameter estimates. This bias results in selection of wrong effects asymptotically, which we can confirm via simulations for more complex settings. By jointly modeling the longitudinal binary data with the dropout process that possibly leads to nonignorable missing data, we are able to correct the bias in estimation and selection. Mixture priors with a point mass at zero are used to facilitate variable selection. We illustrate the proposed approach using a clinical trial for acute ischemic stroke.