共 1467 篇,按发表时间排,新的在前,每页 40 篇。这一页是 2020-04-29 到 2020-04-13。
第 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 页
Many challenging problems in biomedical research rely on understanding how variables are associated with each other and influenced by genetic and environmental factors. Probabilistic graphical models (PGMs) are widely acknowledged as a very natural and formal language to describe relationships among variables and have been extensively used for studying complex diseases and traits. In this work, we propose methods that leverage observational Gaussian family data for learning a decomposition of undirected and directed acyclic PGMs according to the influence of genetic and environmental factors. Many structure learning algorithms are strongly based on a conditional independence test. For independent measurements of normally distributed variables, conditional independence can be tested through standard tests for zero partial correlation. In family data, the assumption of independent measurements does not hold since related individuals are correlated due to mainly genetic factors. Based on univariate polygenic linear mixed models, we propose tests that account for the familial dependence structure and allow us to assess the significance of the partial correlation due to genetic (between‐family) factors and due to other factors, denoted here as environmental (within‐family) factors, separately. Then, we extend standard structure learning algorithms, including the IC/PC and the really fast causal inference (RFCI) algorithms, to Gaussian family data. The algorithms learn the most likely PGM and its decomposition into two components, one explained by genetic factors and the other by environmental factors. The proposed methods are evaluated by simulation studies and applied to the Genetic Analysis Workshop 13 simulated dataset, which captures significant features of the Framingham Heart Study.
The matched case-crossover study design is used in public health, biomedical, and epidemiological research with clustered binary outcomes. Conditional logistic regression is commonly used for analysis because any effects associated with the matching covariates by stratum can be removed. However, some matching covariates often play an important role as effect modifications, causing incorrect statistical testing. The covariates in such studies are often measured with error, so that not accounting for this error can also lead to incorrect inferences for all covariates in the model. However, the methods available for simultaneously evaluating effect modification by matching covariates as well as assessing and characterizing error-in-covariates are limited. In this paper, we propose a flexible omnibus test for testing (1) the significance of a functional association between the clustered binary outcomes and covariates with the measurement error, (2) the existence of effect modifications by matching covariates, and (3) the significance of an interaction effect between the measurement error covariate and other covariates, without specifying the functional forms for these testings. The proposed omnibus test has the flexibility to allow inferences on various hypothesis settings. The advantages of the proposed flexible omnibus test are demonstrated through simulation studies and 1:4 bidirectional matched data analyses from an epidemiology study.
Many longitudinal databases record the occurrence of recurrent events over time. In this article, we propose a new method to estimate the average causal effect of a binary treatment for recurrent event data in the presence of confounders. We propose a doubly robust semiparametric estimator based on a weighted version of the Nelson‐Aalen estimator and a conditional regression estimator under an assumed semiparametric multiplicative rate model for recurrent event data. We show that the proposed doubly robust estimator is consistent and asymptotically normal. In addition, a model diagnostic plot of residuals is presented to assess the adequacy of our proposed semiparametric model. We then evaluate the finite sample behavior of the proposed estimators under a number of simulation scenarios. Finally, we illustrate the proposed methodology via a database of circus artist injuries.
In many multiple testing applications in genetics, the signs of the test statistics provide useful directional information, such as whether genes are potentially up‐ or down‐regulated between two experimental conditions. However, most existing procedures that control the false discovery rate (FDR) are P‐value based and ignore such directional information. We introduce a novel procedure, the signed‐knockoff procedure, to utilize the directional information and control the FDR in finite samples. We demonstrate the power advantage of our procedure through simulation studies and two real applications.
Conventional seamless phase 2/3 design with fixed sample size determination (SSD) has gained its popularity in oncology drug development due to attractive features such as significantly shortening the development timeline, minimizing sample size, as well as early decision making. However, this design is not immune to inaccurate treatment effect assumption when only limited efficacy data are available at study design stage. We propose an innovative seamless phase 2/3 study design with flexible SSD for oncology trials, in which the trial is designed under a distribution of treatment effect instead of one single assumption due to huge uncertainty of treatment effect at design stage and the sample size for end of phase 3 analysis is not predetermined at design stage, but rather dynamically determined based on observed treatment effect at phase 2 portion. Some practical sample size determination rules for end of phase 3 analysis will be discussed. The proposed design can lead to reduced sample size or/and improved power compared with conventional seamless phase 2/3 design with fixed SSD. This innovative study design can be especially useful for programs with aggressive development strategy to expedite the process in delivering efficacious treatment to patients.
Large sample theory of semiparametric models based on maximum likelihood estimation (MLE) with shape constraint on the nonparametric component is well studied. Relatively less attention has been paid to the computational aspect of semiparametric MLE. The computation of semiparametric MLE based on existing approaches such as the expectation‐maximization (EM) algorithm can be computationally prohibitive when the missing rate is high. In this paper, we propose a computational framework for semiparametric MLE based on an inexact block coordinate ascent (BCA) algorithm. We show theoretically that the proposed algorithm converges. This computational framework can be applied to a wide range of data with different structures, such as panel count data, interval‐censored data, and degradation data, among others. Simulation studies demonstrate favorable performance compared with existing algorithms in terms of accuracy and speed. Two data sets are used to illustrate the proposed computational method. We further implement the proposed computational method in R packageBCA1SG, available at CRAN.
In studies based on electronic health records (EHR), the frequency of covariate monitoring can vary by covariate type, across patients, and over time, which can limit the generalizability of inferences about the effects of adaptive treatment strategies. In addition, monitoring is a health intervention in itself with costs and benefits, and stakeholders may be interested in the effect of monitoring when adopting adaptive treatment strategies. This paper demonstrates how to exploit nonsystematic covariate monitoring in EHR‐based studies to both improve the generalizability of causal inferences and to evaluate the health impact of monitoring when evaluating adaptive treatment strategies. Using a real world, EHR‐based, comparative effectiveness research (CER) study of patients with type II diabetes mellitus, we illustrate how the evaluation of joint dynamic treatment and static monitoring interventions can improve CER evidence and describe two alternate estimation approaches based on inverse probability weighting (IPW). First, we demonstrate the poor performance of the standard estimator of the effects of joint treatment‐monitoring interventions, due to a large decrease in data support and concerns over finite‐sample bias from near‐violations of the positivity assumption (PA) for the monitoring process. Second, we detail an alternate IPW estimator using a no direct effect assumption. We demonstrate that this estimator can improve efficiency but at the potential cost of increase in bias from violations of the PA for the treatment process.
In oncology, there is a growing number of therapies given in combination. Recently, several dose-finding designs for Phase I dose-escalation trials for combinations were proposed. The majority of novel designs use a pre-specified parametric model restricting the search of the target combination to a surface of a particular form. In this work, we propose a novel model-free design for combination studies, which is based on the assumption of monotonicity within each agent only. Specifically, we parametrise the ratios between each neighbouring combination by independent Beta distributions. As a result, the design does not require the specification of any particular parametric model or knowledge about increasing orderings of toxicity. We compare the performance of the proposed design to the model-based continual reassessment method for partial ordering and to another model-free alternative, the product of independent beta design. In an extensive simulation study, we show that the proposed design leads to comparable or better proportions of correct selections of the target combination while leading to the same or fewer average number of toxic responses in a trial.
Several studies for the clinical validity of circulating tumor cells (CTCs) in metastatic breast cancer were conducted showing that it is a prognostic biomarker of overall survival. In this work, we consider an individual patient data meta‐analysis for nonmetastatic breast cancer to assess the discrimination of CTCs regarding the risk of death. Data are collected in several centers and present correlated failure times for subjects of the same center. However, although the covariate‐specific time‐dependent receiver operating characteristic (ROC) curve has been widely used for assessing the performance of a biomarker, there is no methodology yet that can handle this specific setting with clustered censored failure times. We propose an estimator for the covariate‐specific time‐dependent ROC curves and area under the ROC curve when clustered failure times are detected. We discuss the assumptions under which the estimators are consistent and their interpretations. We assume a shared frailty model for modeling the effect of the covariates and the biomarker on the outcome in order to account for the cluster effect. A simulation study was conducted and it shows negligible bias for the proposed estimator and a nonparametric one based on inverse probability censoring weighting, while a semiparametric estimator, ignoring the clustering, is markedly biased. Finally, in our application to breast cancer data, the estimation of the covariate‐specific area under the curves illustrates that the CTCs discriminate better patients with inflammatory tumor than patients with noninflammatory tumor, with respect to their risk of death.
In medical research, repeated questionnaire data is often used to measure and model latent variables across time. Through a novel imputation method, a direct comparison is made between latent growth analysis under classical test theory and item response theory, while also including effects of missing item responses. For classical test theory and item response theory, by means of a simulation study the effects of item missingness on latent growth parameter estimates are examined given longitudinal item response data. Several missing data mechanisms and conditions are evaluated in the simulation study. The additional effects of missingness on differences in classical test theory- and item response theory-based latent growth analysis are directly assessed by rescaling the multiple imputations. The multiple imputation method is used to generate latent variable and item scores from the posterior predictive distributions to account for missing item responses in observed multilevel binary response data. It is shown that a multivariate probit model, as a novel imputation model, improves the latent growth analysis, when dealing with missing at random (MAR) in classical test theory. The study also shows that the parameter estimates for the latent growth model using item response theory show less bias and have smaller MSE’s compared to the estimates using classical test theory.
State‐space models (SSMs) are a popular tool for modeling animal abundances. Inference difficulties for simple linear SSMs are well known, particularly in relation to simultaneous estimation of process and observation variances. Several remedies to overcome estimation problems have been studied for relatively simple SSMs, but whether these challenges and proposed remedies apply for nonlinear stage‐structured SSMs, an important class of ecological models, is less well understood. Here we identify improvements for inference about nonlinear stage‐structured SSMs fit with biased sequential life stage data. Theoretical analyses indicate parameter identifiability requires covariates in the state processes. Simulation studies show that plugging in externally estimated observation variances, as opposed to jointly estimating them with other parameters, reduces bias and standard error of estimates. In contrast to previous results for simple linear SSMs, strong confounding between jointly estimated process and observation variance parameters was not found in the models explored here. However, when observation variance was also estimated in the motivating case study, the resulting process variance estimates were implausibly low (near‐zero). As SSMs are used in increasingly complex ways, understanding when inference can be expected to be successful, and what aids it, becomes more important. Our study illustrates (a) the need for relevant process covariates and (b) the benefits of using externally estimated observation variances for inference about nonlinear stage‐structured SSMs.
A control theory perspective on determination of optimal dynamic treatment regimes is considered. The aim is to adapt statistical methodology that has been developed for medical or other biostatistical applications to incorporate powerful control techniques that have been designed for engineering or other technological problems. Data tend to be sparse and noisy in the biostatistical area and interest has tended to be in statistical inference for treatment effects. In engineering fields, experimental data can be more easily obtained and reproduced and interest is more often in performance and stability of proposed controllers rather than modeling and inference per se. We propose that modeling and estimation should be based on standard statistical techniques but subsequent treatment policy should be obtained from robust control. To bring focus, we concentrate on A‐learning methodology as developed in the biostatistical literature and ‐synthesis from control theory. Simulations and two applications demonstrate robustness of the strategy compared to standard A‐learning in the presence of model misspecification or measurement error.
We develop flexible multiparameter regression (MPR) survival models for interval‐censored survival data arising in longitudinal prospective studies and longitudinal randomised controlled clinical trials. A multiparameter Weibull regression survival model, which is wholly parametric, and has nonproportional hazards, is the main focus of the article. We describe the basic model, develop the interval‐censored likelihood, and extend the model to include gamma frailty and a dispersion model. We evaluate the models by means of a simulation study and a detailed reanalysis of data from the Signal Tandmobiel study. The results demonstrate that the MPR model with frailty is computationally efficient and provides an excellent fit to the data.
Consider the problem: given the data pair (x,y) drawn from a population with f*(x)=E[y|x=x], specify a neural network model and run gradient flow on the weights over time until reaching any stationarity. How does ft , the function computed by the neural network at time t, relate to f*, in terms of approximation and representation? What are the provable benefits of the adaptive representation by neural networks compared to the pre-specified fixed basis representation in the classical nonparametric literature? We answer the above questions via a dynamic reproducing kernel Hilbert space (RKHS) approach indexed by the training process of neural networks. Firstly, we show that when reaching any local stationarity, gradient flow learns an adaptive RKHS representation and performs the global least-squares projection onto the adaptive RKHS, simultaneously. Secondly, we prove that as the RKHS is data-adaptive and task-specific, the residual for f* lies in a subspace that is potentially much smaller than the orthogonal complement of the RKHS. The result formalizes the representation and approximation benefits of neural networks. Lastly, we show that the neural network function computed by gradient flow converges to the kernel ridgeless regression with an adaptive kernel, in the limit of vanishing regularization. The adaptive kernel viewpoint provides new angles of studying the approximation, representation, generalization, and optimization advantages of neural networks. *(abstract via OpenAlex)*
Diagnostic test accuracy studies observe the result of a gold standard procedure that defines the presence or absence of a disease and the result of a diagnostic test. They typically report the number of true positives, false positives, true negatives and false negatives. However, diagnostic test outcomes can also be either non-evaluable positives or non-evaluable negatives. We propose a novel model for the meta-analysis of diagnostic studies in the presence of non-evaluable outcomes, which assumes independent multinomial distributions for the true and non-evaluable positives, and, the true and non-evaluable negatives, conditional on the latent sensitivity, specificity, probability of non-evaluable positives and probability of non-evaluable negatives in each study. For the random effects distribution of the latent proportions, we employ a drawable vine copula that can successively model the dependence in the joint tails. Our methodology is demonstrated with an extensive simulation study and applied to data from diagnostic accuracy studies of coronary computed tomography angiography for the detection of coronary artery disease. The comparison of our method with the existing approaches yields findings in the real data application that change the current conclusions.
Summary: A graphical model provides a compact and efficient representation of the association structure in a multivariate distribution by means of a graph. Relevant features of the distribution are represented by vertices, edges and higher-order graphical structures such as cliques or paths. Typically, paths play a central role in these models because they determine the dependence relationships between variables. However, while a theory of path coefficients is available for directed graph models, little research exists on the strength of the association represented by a path in an undirected graph. Essentially, it has been shown that the covariance between two variables can be decomposed into a sum of weights associated with each of the paths connecting the two variables in the corresponding concentration graph. In this context, we consider concentration graph models and provide an extensive analysis of the properties of path weights and their interpretation. Specifically, we give an interpretation of covariance weights through their factorization into a partial covariance and an inflation factor. We then extend the covariance decomposition over the paths of an undirected graph to other measures of association, such as the marginal correlation coefficient and a quantity that we call the inflated correlation. Application of these results is illustrated with an analysis of dietary intake networks.
Valid inference after model selection is currently a very active area of research. The polyhedral method, pioneered by Lee, et al. (2016), allows for valid inference after model selection if the model selection event can be described by polyhedral constraints. In that reference, the method is exemplified by constructing two valid confidence intervals when the Lasso estimator is used to select a model. We here study the length of these intervals. For one of these confidence intervals, which is easier to compute, we find that its expected length is always infinite. For the other of these confidence intervals, whose computation is more demanding, we give a necessary and sufficient condition for its expected length to be infinite. In simulations, we find that this sufficient condition is typically satisfied, unless the selected model includes almost all or almost none of the available regressors. For the distribution of confidence interval length, we find that the $κ$-quantiles behave like $1/(1-κ)$ for $κ$ close to $1$. Our results can also be used to analyze other confidence intervals that are based on the polyhedral method. *(abstract via OpenAlex)*
In 1975, John W. Tukey defined statistical data depth as a function that determines the centrality of an arbitrary point with respect to a data cloud or to a probability measure. During the last decades, this seminal idea of data depth evolved into a powerful tool proving to be useful in various fields of science. Recently, extending the notion of data depth to the functional setting attracted a lot of attention among theoretical and applied statisticians. We go further and suggest a notion of data depth suitable for data represented as curves, or trajectories, which is independent of the parameterization. We show that our curve depth satisfies theoretical requirements of general depth functions that are meaningful for trajectories. We apply our methodology to diffusion tensor brain images and also to pattern recognition of handwritten digits and letters. Supplementary materials for this article are available online. *(abstract via OpenAlex)*
Summary: We study the effect of imperfect training data labels on the performance of classification methods. In a general setting, where the probability that an observation in the training dataset is mislabelled may depend on both the feature vector and the true label, we bound the excess risk of an arbitrary classifier trained with imperfect labels in terms of its excess risk for predicting a noisy label. This reveals conditions under which a classifier trained with imperfect labels remains consistent for classifying uncorrupted test data points. Furthermore, under stronger conditions, we derive detailed asymptotic properties for the popular $k$-nearest neighbour, support vector machine and linear discriminant analysis classifiers. One consequence of these results is that the $k$-nearest neighbour and support vector machine classifiers are robust to imperfect training labels, in the sense that the rate of convergence of the excess risk of these classifiers remains unchanged; in fact, our theoretical and empirical results even show that in some cases, imperfect labels may improve the performance of these methods. The linear discriminant analysis classifier is shown to be typically inconsistent in the presence of label noise unless the prior probabilities of the classes are equal. Our theoretical results are supported by a simulation study.
Summary: Bandwidth choice is crucial in spatial kernel estimation in exploring non-Gaussian complex spatial data. The paper investigates the choice of adaptive and non-adaptive bandwidths for density estimation given data on a spatial lattice. An adaptive bandwidth depends on local data and hence adaptively conforms with local features of the spatial data. We propose a spatial cross-validation (SCV) choice of a global bandwidth. This is done first with a pilot density involved in the expression for the adaptive bandwidth. The optimality of the procedure is established, and it is shown that a non-adaptive bandwidth choice comes out as a special case. Although the cross-validation idea has been popular for choosing a non-adaptive bandwidth in data-driven smoothing of independent and time series data, its theory and application have not been much investigated for spatial data. For the adaptive case, there is little theory even for independent data. Conditions that ensure asymptotic optimality of the SCV-selected bandwidth are derived, actually, also extending time series and independent data optimality results. Further, for the adaptive bandwidth with an estimated pilot density, oracle properties of the resultant density estimator are obtained asymptotically as if the true pilot were known. Numerical simulations show that finite sample performance of the SCV adaptive bandwidth choice works quite well. It outperforms the existing R routines such as the ‘rule of thumb’ and the so-called ‘second-generation’ Sheather–Jones bandwidths for moderate and big data sets. An empirical application to a set of spatial soil data is further implemented with non-Gaussian features significantly identified.
Absent randomization, causal conclusions gain strength if several independent evidence factors concur. We develop a method for constructing evidence factors from several instruments plus a direct comparison of treated and control groups, and we evaluate the methods performance in terms of design sensitivity and simulation. In the application, we consider the effectiveness of Catholic versus public high schools, constructing three evidence factors from three past strategies for studying this question, namely: (i) having nearby access to a Catholic school as an instrument, (ii) being Catholic as an instrument for attending Catholic school, and (iii) a direct comparison of students in Catholic and public high schools. Although these three analyses use the same data, we: (i) construct three essentially independent statistical tests of no effect that require very different assumptions, (ii) study the sensitivity of each test to the assumptions underlying that test, (iii) examine the degree to which independent tests dependent upon different assumptions concur, (iv) pool evidence across independent factors. In the application, we conclude that the ostensible benefit of Catholic education depends critically on the validity of one instrument, and is therefore quite fragile. Supplementary materials for this article, including a standardized description of the materials available for reproducing the work, are available as an online supplement. *(abstract via OpenAlex)*
This article is motivated by the problem of inference on interactions among chemical exposures impacting human health outcomes. Chemicals often co-occur in the environment or in synthetic mixtures and as a result exposure levels can be highly correlated. We propose a latent factor joint model, which includes shared factors in both the predictor and response components while assuming conditional independence. By including a quadratic regression in the latent variables in the response component, we induce flexible dimension reduction in characterizing main effects and interactions. We propose a Bayesian approach to inference under this Factor analysis for INteractions (FIN) framework. Through appropriate modifications of the factor modeling structure, FIN can accommodate higher order interactions. We evaluate the performance using a simulation study and data from the National Health and Nutrition Examination Survey (NHANES). Code is available on GitHub. *(abstract via PubMed)*
Summary: Adaptive enrichment designs involve preplanned rules for modifying enrolment criteria based on accruing data in a randomized trial. We focus on designs where the overall population is partitioned into two predefined subpopulations, e.g. based on a biomarker or risk score measured at baseline. The goal is to learn which populations benefit from an experimental treatment. Two critical components of adaptive enrichment designs are the decision rule for modifying enrolment, and the multiple-testing procedure. We provide a general method for simultaneously optimizing these components for two-stage, adaptive enrichment designs. We minimize the expected sample size under constraints on power and the familywise type I error rate. It is computationally infeasible to solve this optimization problem directly because of its non-convexity. The key to our approach is a novel, discrete representation of this optimization problem as a sparse linear program, which is large but computationally feasible to solve by using modern optimization techniques. We provide an R package that implements our method and is compatible with linear program solvers in several software languages. Our approach produces new, approximately optimal trial designs.
This article addresses the challenges in classifying textual data obtained from open online platforms, which are vulnerable to distortion. Most existing classification methods minimize the overall classification error and may yield an undesirably large Type I error (relevant textual messages are classified as irrelevant), particularly when available data exhibit an asymmetry between relevant and irrelevant information. Data distortion exacerbates this situation and often leads to fallacious prediction. To deal with inestimable data distortion, we propose the use of the Neyman–Pearson (NP) classification paradigm, which minimizes Type II error under a user-specified Type I error constraint. Theoretically, we show that the NP oracle is unaffected by data distortion when the class conditional distributions remain the same. Empirically, we study a case of classifying posts about worker strikes obtained from a leading Chinese microblogging platform, which are frequently prone to extensive, unpredictable and inestimable censorship. We demonstrate that, even though the training and test data are susceptible to different distortion and therefore potentially follow different distributions, our proposed NP methods control the Type I error on test data at the targeted level. The methods and implementation pipeline proposed in our case study are applicable to many other problems involving data distortion. Supplementary materials for this article, including a standardized description of the materials available for reproducing the work, are available as an online supplement. *(abstract via OpenAlex)*
When existing clinical trial data suggest a promising subgroup, we must address the question of how good the selected subgroup really is. The usual statistical inference applied to the selected subgroup, assuming that the subgroup is chosen independent of the data, may lead to an overly optimistic evaluation of the selected subgroup. In this article, we address the issue of selection bias and develop a de-biasing bootstrap inference procedure for the best selected subgroup effect. The proposed inference procedure is model-free, easy to compute, and asymptotically sharp. We demonstrate the merit of our proposed method by reanalyzing the MONET1 trial and show that how the subgroup is selected post hoc should play an important role in any statistical analysis. Supplementary materials for this article are available online. *(abstract via OpenAlex)*
Introduction: In a confirmatory diagnostic accuracy study, sensitivity and specificity are considered as co-primary endpoints. For the sample size calculation, the prevalence of the target population must be taken into account to obtain a representative sample. In this context, a general problem arises. With a low or high prevalence, the study may be overpowered in one subpopulation. One further issue is the correct pre-specification of the true prevalence. With an incorrect assumption about the prevalence, an over- or underestimated sample size will result. Methods: To obtain the desired power independent of the prevalence, a method for an optimal sample size calculation for the comparison of a diagnostic experimental test with a prespecified minimum sensitivity and specificity is proposed. To face the problem of an incorrectly pre-specified prevalence, a blinded one-time re-estimation design of the sample size based on the prevalence and a blinded repeated re-estimation design of the sample size based on the prevalence are evaluated by a simulation study. Both designs are compared to a fixed design and additionally among each other. Results: The type I error rates of both blinded re-estimation designs are not inflated. Their empirical overall power equals the desired theoretical power and both designs offer unbiased estimates of the prevalence. The repeated re-estimation design reveals no advantages concerning the mean squared error of the re-estimated prevalence or sample size compared to the one-time re-estimation design. The appropriate size of the internal pilot study in the one-time re-estimation design is 50% of the initially calculated sample size. Conclusions: A one-time re-estimation design of the prevalence based on the optimal sample size calculation is recommended in single-arm diagnostic accuracy studies.
We provide methods to estimate the confidence interval for the difference between two relative risks. Letting p0, p1, and p2 be the probabilities of an event in three groups (i.e. control, treatment 1, treatment 2), our methods estimate a confidence interval for r = p1/ p0 − p2/ p0. We highlight that our methods can handle small sample sizes, covariates, and study populations from multiple strata. We specifically developed these methods for vaccine trials to estimate the difference between two vaccine efficacies, where VE1 = 1 − p1/ p0, VE2 = 1 − p2/ p0 and r = VE2 − VE1. We showcase our methods by using interim data from one of these trials to suggest that one dose of the human papillomavirus vaccine may be as efficacious as two doses of the vaccine.
Clinical and health care processes are often summarised through sequences of ordinal data describing patient’s state over time. Identifying patterns in these sequences can provide valuable insights into patient progression trajectories for the purposes of clinical monitoring and quality assurance. However, both the variation in the length of each sequence and the ordinal nature of observable states present challenges to pattern identification. In this paper, we address these challenges by presenting a novel measure of dissimilarity for comparing two or more variable-length ordinal sequences that can be used in conjunction with conventional clustering methods to identify patterns in patient progression trajectories. We provide practical guidance on how this can be achieved, and demonstrate it in the context of identifying patterns in post-stoke recovery trajectories.
Currently, methods for conducting multiple treatment propensity scoring in the presence of high‐dimensional covariate spaces that result from “big data” are lacking—the most prominent method relies on inverse probability treatment weighting (IPTW). However, IPTW only utilizes one element of the generalized propensity score (GPS) vector, which can lead to a loss of information and inadequate covariate balance in the presence of multiple treatments. This limitation motivates the development of a novel propensity score method that uses the entire GPS vector to establish a scalar balancing score that, when adjusted for, achieves covariate balance in the presence of potentially high‐dimensional covariates. Specifically, the generalized propensity score cumulative distribution function (GPS‐CDF) method is introduced. A one‐parameter power function fits the CDF of the GPS vector and a resulting scalar balancing score is used for matching and/or stratification. Simulation results show superior performance of the new method compared to IPTW both in achieving covariate balance and estimating average treatment effects in the presence of multiple treatments. The proposed approach is applied to a study derived from electronic medical records to determine the causal relationship between three different vasopressors and mortality in patients with non‐traumatic aneurysmal subarachnoid hemorrhage. Results suggest that the GPS‐CDF method performs well when applied to large observational studies with multiple treatments that have large covariate spaces.
Summary: Canonical correlation analysis investigates linear relationships between two sets of variables, but it often works poorly on modern datasets because of high dimensionality and mixed data types such as continuous, binary and zero-inflated. To overcome these challenges, we propose a semiparametric approach to sparse canonical correlation analysis based on the Gaussian copula. The main result of this paper is a truncated latent Gaussian copula model for data with excess zeros, which allows us to derive a rank-based estimator of the latent correlation matrix for mixed variable types without estimation of marginal transformation functions. The resulting canonical correlation analysis method works well in high-dimensional settings, as demonstrated via numerical studies, and when applied to the analysis of association between gene expression and microRNA data from breast cancer patients.
Summary: We provide a new computationally efficient class of estimators for risk minimization. We show that these estimators are robust for general statistical models, under varied robustness settings, including in the classical Huber ε-contamination model, and in heavy-tailed settings. Our workhorse is a novel robust variant of gradient descent, and we provide conditions under which our gradient descent variant provides accurate estimators in a general convex risk minimization problem. We provide specific consequences of our theory for linear regression and logistic regression and for canonical parameter estimation in an exponential family. These results provide some of the first computationally tractable and provably robust estimators for these canonical statistical models. Finally, we study the empirical performance of our proposed methods on synthetic and real data sets, and we find that our methods convincingly outperform a variety of baselines.
Methods for the evaluation of the predictive accuracy of biomarkers with respect to survival outcomes subject to right censoring have been discussed extensively in the literature. In cancer and other diseases, survival outcomes are commonly subject to interval censoring by design or due to the follow up schema. In this article, we present an estimator for the area under the time‐dependent receiver operating characteristic (ROC) curve for interval censored data based on a nonparametric sieve maximum likelihood approach. We establish the asymptotic properties of the proposed estimator and illustrate its finite‐sample properties using a simulation study. The application of our method is illustrated using data from a cancer clinical study. An open‐source R package to implement the proposed method is available on Comprehensive R Archive Network.
Publication bias frequently appears in meta-analyses when the included studies’ results (e.g., p-values) influence the studies’ publication processes. Some unfavorable studies may be suppressed from publication, so the meta-analytic results may be biased toward an artificially favorable direction. Many statistical tests have been proposed to detect publication bias in recent two decades. However, they often make dramatically different assumptions about the cause of publication bias; therefore, they are usually powerful only in certain cases that support their particular assumptions, while their powers may be fairly low in many other cases. Although several simulation studies have been carried out to compare different tests’ powers under various situations, it is typically infeasible to justify the exact mechanism of publication bias in a real-world meta-analysis and thus select the corresponding optimal publication bias test. We introduce a hybrid test for publication bias by synthesizing various tests and incorporating their benefits, so that it maintains relatively high powers across various mechanisms of publication bias. The superior performance of the proposed hybrid test is illustrated using simulation studies and three real-world meta-analyses with different effect sizes. It is compared with many existing methods, including the commonly used regression and rank tests, and the trim-and-fill method.
Summary: Meta-analysis has become a powerful tool for improving inference by gathering evidence from multiple sources. It pools summary-level data from different studies to improve estimation efficiency with the assumption that all participating studies are analysed under the same statistical model. It is challenging to integrate external summary data calculated from different models with a newly conducted internal study in which individual-level data are collected. We develop a novel statistical inference framework that can effectively synthesize internal and external data for the integrative analysis. The new framework is versatile enough to assimilate various types of summary data from multiple sources. We establish asymptotic properties for the proposed procedure and prove that the new estimate is theoretically more efficient than the internal data based maximum likelihood estimate, as well as a recently developed constrained maximum likelihood approach that incorporates the external information. We illustrate an application of our method by evaluating cervical cancer risk using data from a large cervical screening program.
Summary: We consider scenarios in which the likelihood function for a semiparametric regression model factors into separate components, with an efficient estimator of the regression parameter available for each component. An optimal weighted combination of the component estimators, named an ensemble estimator, may be employed as an overall estimate of the regression parameter, and may be fully efficient under uncorrelatedness conditions. This approach is useful when the full likelihood function may be difficult to maximize, but the components are easy to maximize. It covers settings where the nuisance parameter may be estimated at different rates in the component likelihoods. As a motivating example we consider proportional hazards regression with prospective doubly censored data, in which the likelihood factors into a current status data likelihood and a left-truncated right-censored data likelihood. Variable selection is important in such regression modelling, but the applicability of existing techniques is unclear in the ensemble approach. We propose ensemble variable selection using the least squares approximation technique on the unpenalized ensemble estimator, followed by ensemble re-estimation under the selected model. The resulting estimator has the oracle property such that the set of nonzero parameters is successfully recovered and the semiparametric efficiency bound is achieved for this parameter set. Simulations show that the proposed method performs well relative to alternative approaches. Analysis of an AIDS cohort study illustrates the practical utility of the method.
The development of molecular diagnostic tools to achieve individualized medicine requires identifying predictive biomarkers associated with subgroups of individuals who might receive beneficial or harmful effects from different available treatments. However, due to the large number of candidate biomarkers in the large-scale genetic and molecular studies, and complex relationships among clinical outcome, biomarkers, and treatments, the ordinary statistical tests for the interactions between treatments and covariates have difficulties from their limited statistical powers. In this paper, we propose an efficient method for detecting predictive biomarkers. We employ weighted loss functions of Chen et al. to directly estimate individual treatment scores and propose synthetic posterior inference for effect sizes of biomarkers. We develop an empirical Bayes approach, namely, we estimate unknown hyperparameters in the prior distribution based on data. We then provide efficient screening methods for the candidate biomarkers via optimal discovery procedure with adequate control of false discovery rate. The proposed method is demonstrated in simulation studies and an application to a breast cancer clinical study in which the proposed method was shown to detect the much larger numbers of significant biomarkers than existing standard methods.
Characterization of HIV viral rebound after the discontinuation of antiretroviral therapy is central to HIV cure research. We propose a parametric nonlinear mixed effects model for the viral rebound trajectory, which often has a rapid rise to a peak value followed by a decrease to a viral load set point. We choose a flexible functional form that captures the shapes of viral rebound trajectories and can also provide biological insights regarding the rebound process. Each parameter can incorporate a random effect to allow for variation in parameters across individuals. Key features of viral rebound trajectories such as viral set points are represented by the parameters in the model, which facilitates assessment of intervention effects and identification of important pretreatment interruption predictors for these features. We employ a stochastic expectation‐maximization (StEM) algorithm to incorporate HIV‐1 RNA values that are below the lower limit of assay quantification. We evaluate the performance of our model in simulation studies and apply the proposed model to longitudinal HIV‐1 viral load data from five AIDS Clinical Trials Group treatment interruption studies.
In genomic analysis, it is significant though challenging to identify markers associated with cancer outcomes or phenotypes. Based on the biological mechanisms of cancers and the characteristics of datasets, we propose a novel integrative interaction approach under a semiparametric model, in which genetic and environmental factors are included as the parametric and nonparametric components, respectively. The goal of this approach is to identify the genetic factors and gene–gene interactions associated with cancer outcomes, while estimating the nonlinear effects of environmental factors. The proposed approach is based on the threshold gradient-directed regularisation technique. Simulation studies indicate that the proposed approach outperforms alternative methods at identifying the main effects and interactions, and has favourable estimation and prediction accuracy. We analysed non-small-cell lung carcinoma datasets from the Cancer Genome Atlas, and the results demonstrate that the proposed approach can identify markers with important implications and that it performs favourably in terms of prediction accuracy, identification stability, and computation cost.
第 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 页