Annals of Applied Statistics最新文献

It is of substantial interest to study health disparity associations with COVID-19 death rates. Although high-quality individual-level COVID-19 epidemiological data have been difficult to collect on a national scale, all United States (U.S.) counties have reported total COVID-19 death counts. A standard ecological analysis would then regress county total death counts by county-level covariates such as age, sex, and race percentages. However, such an analysis is limited by ecological bias and fallacy in which estimated county-level associations are different from individual-level associations. Fortunately, state-level age, sex, and race specific COVID-19 death counts are also available for all U.S. states, so this information can be integrated with county-level data for more informative ecological analyses. We propose an approximate log-linear random effects model to jointly model county-level total death counts and state-level age, sex, and race specific death counts. We then develop a penalized composite log-likelihood method for parameter estimation and perform simulation studies to evaluate our proposed approach. Lastly, we analyze COVID-19 death data from the entire U.S., show how incorporating state-level counts can prevent ecological bias and fallacy, and illustrate the heterogeneity in health disparity associations across different U.S. states.

Sepsis is a life-threatening condition caused by a dysregulated host response to infection. Recently, researchers have hypothesized that sepsis consists of a heterogeneous spectrum of distinct subtypes, motivating several studies to identify clusters of sepsis patients that correspond to subtypes, with the long-term goal of using these clusters to design subtype-specific treatments. Therefore, clinicians rely on clusters having a concrete medical interpretation, usually corresponding to clinically meaningful regions of the sample space that have a concrete implication to practitioners. In this article, we propose Clustering Around Meaningful Regions (CLAMR), a Bayesian clustering approach that explicitly models the medical interpretation of each cluster center. CLAMR favors clusterings that can be summarized via meaningful feature values, leading to medically significant sepsis patient clusters. We also provide details on measuring the effect of each feature on the clustering using Bayesian hypothesis tests, so one can assess what features are relevant for cluster interpretation. Our focus is on clustering sepsis patients from Moshi, Tanzania, where patients are younger and the prevalence of HIV infection is higher than in previous sepsis subtyping cohorts.

Observational zero-inflated count data arise in a wide range of areas such as genomics. One of the common research questions is to identify causal relationships by learning the structure of a sparse directed acyclic graph (DAG). While structure learning of DAGs has been an active research area, existing methods do not adequately account for excessive zeros and therefore are not suitable for modeling zero-inflated count data. Moreover, it is often interesting to study differences in the causal networks for data collected from two experimental groups (control vs treatment). To explicitly account for zero-inflation and identify differential causal networks, we propose a novel Bayesian differential zero-inflated negative binomial DAG (DAG0) model. We prove that the causal relationships under the proposed DAG0 are fully identifiable from purely observational, cross-sectional data, using a general proof technique that is applicable beyond the proposed model. Bayesian inference based on parallel-tempered Markov chain Monte Carlo is developed to efficiently explore the multi-modal posterior landscape. We demonstrate the utility of the proposed DAG0 by comparing it with state-of-the-art alternative methods through extensive simulations. An application in a single-cell RNA-sequencing dataset generated under two experimental groups finds some interesting results that appear to be consistent with existing knowledge. A user-friendly R package that implements DAG0 is available at https://github.com/junsoukchoi/BayesDAG0.git.

To investigate causal impacts, many researchers use controlled pre-post designs that compare over-time differences between a population exposed to a policy change and an unexposed comparison group. However, researchers using these designs often disagree about the "correct" specification of the causal model, perhaps most notably in analyses to identify the effects of gun policies on crime. To help settle these model specification debates, we propose a general identification framework that unifies a variety of models researchers use in practice. In this framework, which nests "brand name" designs like Difference-in-Differences as special cases, we use models to predict untreated outcomes and then correct the treated group's predictions using the comparison group's observed prediction errors. Our point identifying assumption is that treated and comparison groups would have equal prediction errors (in expectation) under no treatment. To choose among candidate models, we propose a data-driven procedure based on models' robustness to violations of this point identifying assumption. Our selection procedure averages over candidate models, weighting by each model's posterior probability of being the most robust given its differential average prediction errors in the pre-period. This approach offers a way out of debates over the "correct" model by choosing on robustness instead and has the desirable property of being feasible in the "locked box" of pre-intervention data only. We apply our methodology to the gun policy debate, focusing specifically on Missouri's 2007 repeal of its permit-to-purchase law, and provide an R package (apm) for implementation.

Surrogate selection is an experimental design that without sequencing any DNA can restrict a sample of cells to those carrying certain genomic mutations. In immunological disease studies, this design may provide a relatively easy approach to enrich a lymphocyte sample with cells relevant to the disease response because the emergence of neutral mutations associates with the proliferation history of clonal subpopulations. A statistical analysis of clonotype sizes provides a structured, quantitative perspective on this useful property of surrogate selection. Our model specification couples within-clonotype birth-death processes with an exchangeable model across clonotypes. Beyond enrichment questions about the surrogate selection design, our framework enables a study of sampling properties of elementary sample diversity statistics; it also points to new statistics that may usefully measure the burden of somatic genomic alterations associated with clonal expansion. We examine statistical properties of immunological samples governed by the coupled model specification, and we illustrate calculations in surrogate selection studies of melanoma and in single-cell genomic studies of T cell repertoires.

Contrastive dimension reduction methods have been developed for case-control study data to identify variation that is enriched in the foreground (case) data $X$ relative to the background (control) data $Y$ . Here we develop contrastive regression for the setting where there is a response variable $r$ associated with each foreground observation. This situation occurs frequently when, for example, the unaffected controls do not have a disease grade or intervention dosage, but the affected cases have a disease grade or intervention dosage, as in autism severity, solid tumors stages, polyp sizes, or warfarin dosages. Our contrastive regression model captures shared low-dimensional variation between the predictors in the case and control groups and then explains the case-specific response variables through the variance that remains in the predictors after shared variation is removed. We show that, in one single-cell RNA sequencing dataset on cellular differentiation in chronic rhinosinusitis with and without nasal polyps and in another single-nucleus RNA sequencing dataset on autism severity in postmortem brain samples from donors with and without autism, our contrastive linear regression performs feature ranking and identifies biologically-informative predictors associated with response that cannot be identified using other approaches.

The diagnosis and treatment of cancer can evoke a variety of adverse emotions. Online health communities (OHCs) provide a safe platform for cancer patients and those closely related to express emotions without fear of judgement or stigma. In the literature, linguistic analysis of OHCs is usually limited to a single disease and based on methods with various technical limitations. In this article we analyze posts from September 2003 to September 2022 on eight cancers that are publicly available at the American Cancer Society's Cancer Survivors Network (CSN). We propose a novel network analysis technique based on low-rank matrices. The proposed approach decomposes the emotional expression semantic networks into an across-cancer time-independent component (which describes the "baseline" that is shared by multiple cancers), a cancer-specific time-independent component (which describes cancer-specific properties), and an across-cancer time-dependent component (which accommodates temporal effects on multiple cancer communities). For the second and third components, respectively, we consider a novel clustering structure and a change point structure. A penalization approach is developed, and its theoretical and computational properties are carefully established. The analysis of the CSN data leads to sensible networks and deeper insights into emotions for cancer overall and specific cancer types.

Prediction of time-to-event data often suffers from rare event rates, small sample sizes, high dimensionality, and low signal-to-noise ratios. Incorporating published prediction models from external large-scale studies is expected to improve the performance of prognosis prediction from internal individual-level data. However, existing integration approaches typically assume that the underlying distributions of the external and internal data sources are similar, which is often invalid. To account for challenges, including heterogeneity, data sharing, and privacy constraints, we propose a failure time integration procedure, which utilizes a discrete hazard-based Kullback-Leibler discriminatory information measuring the discrepancy between the external models and the internal dataset. The asymptotic properties and simulation results show the advantage of the proposed method compared to those solely based on internal data. We apply the proposed method to improve prediction performance on a kidney transplant dataset from a local hospital by integrating this small-sized dataset with a published survival model obtained from the national transplant registry.

Semi-continuous data frequently arise in clinical practice. For example, while many surgical patients still suffer from varying degrees of acute postoperative pain (POP) sometime after surgery (i.e., POP score > 0), others experience none (i.e., POP score = 0), indicating the existence of two distinct data processes at play. Existing parametric or semi-parametric two-part modeling methods for this type of semi-continuous data can fail to appropriately model the two underlying data processes as such methods rely heavily on (generalized) linear additive assumptions. However, many factors may interact to jointly influence the experience of POP non-additively and non-linearly. Motivated by this challenge and inspired by the flexibility of deep neural networks (DNN) to accurately approximate complex functions universally, we derive a DNN-based two-part model by adapting the conventional DNN methods with two additional components: a bootstrapping procedure along with a filtering algorithm to boost the stability of the conventional DNN, an approach we denote as sDNN. To improve the interpretability and transparency of sDNN, we further derive a feature importance testing procedure to identify important features associated with the outcome measurements of the two data processes, denoting this approach fsDNN. We show that fsDNN not only offers a statistical inference procedure for each feature under complex association but also that using the identified features can further improve the predictive performance of sDNN. The proposed sDNN- and fsDNN-based two-part models are applied to the analysis of real data from a POP study, in which application they clearly demonstrate advantages over the existing parametric and semi-parametric two-part models. Further, we conduct extensive numerical studies and draw comparisons with other machine learning methods to demonstrate that sDNN and fsDNN consistently outperform the existing two-part models and frequently used machine learning methods regardless of the data complexity. An R package implementing the proposed methods has been developed and is available in the Supplementary Material (Zou et al, 2025) and is also deposited on GitHub (https://github.com/BZou-lab/fsDNN).

Electronic medical records (EMR) data contain rich information that can facilitate health-related studies but is collected primarily for purposes other than research. For recurrent events, EMR data often do not record event times or counts but only contain intermittently assessed and censored observations (i.e. upper and/or lower bounds for counts in a time interval) at uncontrolled times. This can result in non-contiguous or overlapping assessment intervals with censored event counts. Existing methods for analyzing intermittently assessed recurrent events assume disjoint assessment intervals with known counts (interval count data) due to a focus on prospective studies with controlled assessment times. We propose a Bayesian data augmentation method to analyze the complicated assessments in EMR data for recurrent events. Within a Gibbs sampler, event times are imputed by generating sets of event times from non-homogeneous Poisson processes and rejecting proposed sets that are incompatible with constraints imposed by assessment data. Based on the independent increments property of Poisson processes, we implement three techniques to speed up this rejection sampling imputation method for large EMR datasets: independent sampling by partitioning, truncated generation, and sequential sampling. In a simulation study we show our method accurately estimates parameters of log-linear Poisson process intensities. Although the proposed method can be applied generally to EMR data of recurrent events, our study is specifically motivated by identifying risk factors for falls due to cancer treatment and its supportive medications. We used the proposed method to analyze an EMR dataset comprising 5501 patients treated for breast cancer. Our analysis provides evidence supporting associations between certain risk factors (including classes of medications) and risk of falls.