Journal of Data Science

Analyzing the gene-environment interaction (GEI) is crucial for understanding the etiology of many complex traits. Among various types of study designs, case-control studies are popular for analyzing gene-environment interactions due to their efficiency in collecting covariate information. Extensive literature explores efficient estimation under various assumptions made about the relationship between genetic and environmental variables. In this paper, we comprehensively review the methods based on or related to the retrospective likelihood, including the methods based on the hypothetical population concept, which has been largely overlooked in GEI research in the past decade. Furthermore, we establish the methodological connection between these two groups of methods by deriving a new estimator from both the retrospective likelihood and the hypothetical population perspectives. The validity of the derivation is demonstrated through numerical studies.

Predicting the timing and occurrence of events is a major focus of data science applications, especially in the context of biomedical research. Performance for models estimating these outcomes, often referred to as time-to-event or survival outcomes, is frequently summarized using measures of discrimination, in particular time-dependent AUC and concordance. Many estimators for these quantities have been proposed which can be broadly categorized as either semi-parametric estimators or non-parametric estimators. In this paper, we review the mathematical construction of the two classes of estimators and compare their behavior. Importantly, we identify a previously unknown feature of the class of semi-parametric estimators that can result in vastly overoptimistic out-of-sample estimation of discriminative performance in common applied tasks. Although these semi-parametric estimators are popular in practice, the phenomenon we identify here suggests that this class of estimators may be inappropriate for use in model assessment and selection based on out-of-sample evaluation criteria. This is due to the semi-parametric estimators’ bias in favor of models that are overfit when using out-of-sample prediction criteria (e.g. cross-validation). Non-parametric estimators, which do not exhibit this behavior, are highly variable for local discrimination. We propose to address the high variability problem through penalized regression splines smoothing. The behavior of various estimators of time-dependent AUC and concordance are illustrated via a simulation study using two different mechanisms that produce overoptimistic out-of-sample estimates using semi-parametric estimators. Estimators are further compared using a case study using data from the National Health and Nutrition Examination Survey (NHANES) 2011–2014.

When comparing two survival curves, three tests are widely used: the Cox proportional hazards test, the logrank test, and the Wilcoxon test. Despite their popularity in survival data analysis, there is no clear clinical interpretation especially when the proportional hazard assumption is not valid. Meanwhile, the restricted mean survival time (RMST) offers an intuitive and clinically meaningful interpretation. We compare these four tests with regards to statistical power under many configurations (e.g., proportional hazard, early benefit, delayed benefit, and crossing survivals) with data simulated from the Weibull distributions. We then use an example from a lung cancer trial to compare their required sample sizes. As expected, the CoxPH test is more powerful than others when the PH assumption is valid. The Wilcoxon test is often preferable when there is a decreasing trajectory in the event rate as time goes. The RMST test is much more powerful than others when a new treatment has early benefit. The recommended test(s) under each configuration are suggested in this article.

The rapidly expanding field of metabolomics presents an invaluable resource for understanding the associations between metabolites and various diseases. However, the high dimensionality, presence of missing values, and measurement errors associated with metabolomics data can present challenges in developing reliable and reproducible approaches for disease association studies. Therefore, there is a compelling need for robust statistical analyses that can navigate these complexities to achieve reliable and reproducible disease association studies. In this paper, we construct algorithms to perform variable selection for noisy data and control the False Discovery Rate when selecting mutual metabolomic predictors for multiple disease outcomes. We illustrate the versatility and performance of this procedure in a variety of scenarios, dealing with missing data and measurement errors. As a specific application of this novel methodology, we target two of the most prevalent cancers among US women: breast cancer and colorectal cancer. By applying our method to the Women’s Health Initiative data, we successfully identify metabolites that are associated with either or both of these cancers, demonstrating the practical utility and potential of our method in identifying consistent risk factors and understanding shared mechanisms between diseases.

In causal mediation analyses, of interest are the direct or indirect pathways from exposure to an outcome variable. For observation studies, massive baseline characteristics are collected as potential confounders to mitigate selection bias, possibly approaching or exceeding the sample size. Accordingly, flexible machine learning approaches are promising in filtering a subset of relevant confounders, along with estimation using the efficient influence function to avoid overfitting. Among various confounding selection strategies, two attract growing attention. One is the popular debiased, or double machine learning (DML), and another is the penalized partial correlation via fitting a Gaussian graphical network model between the confounders and the response variable. Nonetheless, for causal mediation analyses when encountering high-dimensional confounders, there is a gap in determining the best strategy for confounding selection. Therefore, we exemplify a motivating study on the human microbiome, where the dimensions of mediator and confounders approach or exceed the sample size to compare possible combinations of confounding selection methods. By deriving the multiply robust causal direct and indirect effects across various hypotheses, our comprehensive illustrations offer methodological implications on how the confounding selection impacts the final causal target parameter estimation while generating causality insights in demystifying the “gut-brain axis”. Our results highlighted the practicality and necessity of the discussed methods, which not only guide real-world applications for practitioners but also motivate future advancements for this crucial topic in the era of big data.

After the onset of the COVID-19 pandemic, scientific interest in coronaviruses endemic in animal populations has increased dramatically. However, investigating the prevalence of disease in animal populations across the landscape, which requires finding and capturing animals can be difficult. Spatial random sampling over a grid could be extremely inefficient because animals can be hard to locate, and the total number of samples may be small. Alternatively, preferential sampling, using existing knowledge to inform sample location, can guarantee larger numbers of samples, but estimates derived from this sampling scheme may exhibit bias if there is a relationship between higher probability sampling locations and the disease prevalence. Sample specimens are commonly grouped and tested in pools which can also be an added challenge when combined with preferential sampling. Here we present a Bayesian method for estimating disease prevalence with preferential sampling in pooled presence-absence data motivated by estimating factors related to coronavirus infection among Mexican free-tailed bats (Tadarida brasiliensis) in California. We demonstrate the efficacy of our approach in a simulation study, where a naive model, not accounting for preferential sampling, returns biased estimates of parameter values; however, our model returns unbiased results regardless of the degree of preferential sampling. Our model framework is then applied to data from California to estimate factors related to coronavirus prevalence. After accounting for preferential sampling impacts, our model suggests small prevalence differences between male and female bats.

Estimating healthcare expenditures is important for policymakers and clinicians. The expenditure of patients facing a life-threatening illness can often be segmented into four distinct phases: diagnosis, treatment, stable, and terminal phases. The diagnosis phase encompasses healthcare expenses incurred prior to the disease diagnosis, attributed to frequent healthcare visits and diagnostic tests. The second phase, following diagnosis, typically witnesses high expenditure due to various treatments, gradually tapering off over time and stabilizing into a stable phase, and eventually to a terminal phase. In this project, we introduce a pre-disease phase preceding the diagnosis phase, serving as a baseline for healthcare expenditure, and thus propose a five-phase to evaluate the healthcare expenditures. We use a piecewise linear model with three population-level change points and $4p$ subject-level parameters to capture expenditure trajectories and identify transitions between phases, where p is the number of covariates. To estimate the model’s coefficients, we apply generalized estimating equations, while a grid-search approach is used to estimate the change-point parameters by minimizing the residual sum of squares. In our analysis of expenditures for stages I–III pancreatic cancer patients using the SEER-Medicare database, we find that the diagnostic phase begins one month before diagnosis, followed by an initial treatment phase lasting three months. The stable phase continues until eight months before death, at which point the terminal phase begins, marked by a renewed increase in expenditures.