The Semiparametric Case‐Only Estimator

Summary We propose a semiparametric case‐only estimator of multiplicative gene–environment or gene–gene interactions, under the assumption of conditional independence of the two factors given a vector of potential confounding variables. Our estimator yields valid inferences on the interaction function if either but not necessarily both of two unknown baseline functions of the confounders is correctly modeled. Furthermore, when both models are correct, our estimator has the smallest possible asymptotic variance for estimating the interaction parameter in a semiparametric model that assumes that at least one but not necessarily both baseline models are correct.     

A simple Bayesian analysis of misclassified binary data with a validation substudy

A two-stage Bayesian method is presented for analyzing case-control studies in which a binary variable is sometimes measured with error but the correct values of the variable are known for a random subset of the study group. The first stage of the method is analytically tractable and MCMC methods are used for the second stage. The posterior distribution from the first stage becomes the prior distribution for the second stage, thus transferring all relevant information between the stages. The method makes few distributional assumptions and requires no asymptotic approximations. It is computationally fast and can be run using standard software. It is applied to two data sets that have been analyzed by other methods, and results are compared.     

Matrix methods for estimating odds ratios with misclassified exposure data: extensions and comparisons

Misclassification of exposure variables is a common problem in epidemiologic studies. This paper compares the matrix method (Barron, 1977, Biometrics 33, 414-418; Greenland, 1988a, Statistics in Medicine 7, 745-757) and the inverse matrix method (Marshall, 1990, Journal of Clinical Epidemiology 43, 941-947) to the maximum likelihood estimator (MLE) that corrects the odds ratio for bias due to a misclassified binary covariate. Under the assumption of differential misclassification, the inverse matrix method is always more efficient than the matrix method; however, the efficiency depends strongly on the values of the sensitivity, specificity, baseline probability of exposure, the odds ratio, case-control ratio, and validation sampling fraction. In a study on sudden infant death syndrome (SIDS), an estimate of the asymptotic relative efficiency (ARE) of the inverse matrix estimate was 0.99, while the matrix method's ARE was 0.19. Under nondifferential misclassification, neither the matrix nor the inverse matrix estimator is uniformly more efficient than the other; the efficiencies again depend on the underlying parameters. In the SIDS data, the MLE was more efficient than the matrix method (ARE = 0.39). In a study investigating the effect of vitamin A intake on the incidence of breast cancer, the MLE was more efficient than the matrix method (ARE = 0.75).     

Complementary Log–Log Regression for the Estimation of Covariate‐Adjusted Prevalence Ratios in the Analysis of Data from Cross‐Sectional Studies

We assessed complementary log–log (CLL) regression as an alternative statistical model for estimating multivariable‐adjusted prevalence ratios (PR) and their confidence intervals. Using the delta method, we derived an expression for approximating the variance of the PR estimated using CLL regression. Then, using simulated data, we examined the performance of CLL regression in terms of the accuracy of the PR estimates, the width of the confidence intervals, and the empirical coverage probability, and compared it with results obtained from log–binomial regression and stratified Mantel–Haenszel analysis. Within the range of values of our simulated data, CLL regression performed well, with only slight bias of point estimates of the PR and good confidence interval coverage. In addition, and importantly, the computational algorithm did not have the convergence problems occasionally exhibited by log–binomial regression. The technique is easy to implement in SAS (SAS Institute, Cary, NC), and it does not have the theoretical and practical issues associated with competing approaches. CLL regression is an alternative method of binomial regression that warrants further assessment.     

Combined analysis of transcriptome and proteome data as a tool for the identification of candidate biomarkers in renal cell carcinoma

Results obtained from expression profilings of renal cell carcinoma using different “ome”‐based approaches and comprehensive data analysis demonstrated that proteome‐based technologies and cDNA microarray analyses complement each other during the discovery phase for disease‐related candidate biomarkers. The integration of the respective data revealed the uniqueness and complementarities of the different technologies. While comparative cDNA microarray analyses though restricted to up‐regulated targets largely revealed genes involved in controlling gene/protein expression (19%) and signal transduction processes (13%), proteomics/PROTEOMEX‐defined candidate biomarkers include enzymes of the cellular metabolism (36%), transport proteins (12%), and cell motility/structural molecules (10%). Candidate biomarkers defined by proteomics and PROTEOMEX are frequently shared, whereas the sharing rate between cDNA microarray and proteome‐based profilings is limited. Putative candidate biomarkers provide insights into their cellular (dys)function and their diagnostic/prognostic value but still warrant further validation in larger patient numbers. Based on the fact that merely three candidate biomarkers were shared by all applied technologies, namely annexin A4, tubulin α‐1A chain, and ubiquitin carboxyl‐terminal hydrolase L1, the analysis at a single hierarchical level of biological regulation seems to provide only limited results thus emphasizing the importance and benefit of performing rather combinatorial screenings which can complement the standard clinical predictors.     

Aspects of the design and analysis of high-dimensional SNP studies for disease risk estimation

The state of readiness for high-dimensional single nucleotide polymorphism (SNP) epidemiologic association studies is described, as background for a discussion of statistical aspects of case-control study design and analysis. Specifically, the important role that multistage designs can play in the elimination of false-positive associations and in the control of study costs will be noted. Also, the trade-offs associated with using pooled DNA at early design stages for additional important cost reductions will be discussed in some detail. An odds ratio approach to relating SNP alleles to disease risk using pooled DNA will be proposed, in conjunction with a simple empirical variance estimator, based on comparisons among log-odds ratio estimators from distinct pairs of case and control pools. Simulation studies will be presented to evaluate the moderate sample size properties of such multistage designs and estimation procedures. The design of an ongoing three-stage study in the Women's Health Initiative to relate 250,000 SNPs to the risk of coronary heart disease, stroke, and breast cancer will provide illustration, and will be used to motivate the choice of simulation configurations.     

The unique peptidome: Taxon‐specific tryptic peptides as biomarkers for targeted metaproteomics

The Unique Peptide Finder ( http://unipept.ugent.be/peptidefinder ) is an interactive web application to quickly hunt for tryptic peptides that are unique to a particular species, genus, or any other taxon. Biodiversity within the target taxon is represented by a set of proteomes selected from a monthly updated list of complete and nonredundant UniProt proteomes, supplemented with proprietary proteomes loaded into persistent local browser storage. The software computes and visualizes pan and core peptidomes as unions and intersections of tryptic peptides occurring in the selected proteomes. In addition, it also computes and displays unique peptidomes as the set of all tryptic peptides that occur in all selected proteomes but not in any UniProt record not assigned to the target taxon. As a result, the unique peptides can serve as robust biomarkers for the target taxon, for example, in targeted metaproteomics studies. Computations are extremely fast since they are underpinned by the Unipept database, the lowest common ancestor algorithm implemented in Unipept and modern web technologies that facilitate in‐browser data storage and parallel processing.     

Models for analysing species’ presence/absence data at two time points

Species’ presence/absence at two time points is a very common form of ecological data. It is the simplest type of longitudinal study and has fundamental applications in ecological succession, environmental monitoring, and climate change scenarios. Despite its widespread commonality the use of statistical regression to analyse such data has been wanting. We propose the use of the bivariate odds-ratio model to analyse these data. Seldomly used in ecology, it is argued as being suitable, especially within a constrained ordination framework. In particular, this paper presents the constrained ordination-odds ratio framework as a potentially important key in understanding the underlying processes of niche theory dynamics, e.g., local extinction and colonization probabilities can be described in terms of it. Some of the mathematical and statistical challenges associated with more ambitious extensions are highlighted. As examples, with an underlying Poisson abundance model, a complementary log-log link for the marginal probabilities is shown to be more appropriate. We then develop this model based on the zero-inflated Poisson distribution since excess absences relative to a Poisson distribution is frequent in practice. Two vegetation data sets are used for illustrative purposes.     

Behavioral testing of antidepressant compounds: an analysis of crossover design for correlated binary data

The differential reinforcement of low-rate 72 seconds schedule (DRL-72) is a standard behavioral test procedure for screening potential antidepressant compounds. The protocol for the DRL-72 experiment, proposed by Evenden et al. (1993), consists of using a crossover design for the experiment and one-way ANOVA for the statistical analysis. In this paper we discuss the choice of several crossover designs for the DRL-72 experiment and propose to estimate the treatment effects using either generalized linear mixed models (GLMM) or generalized estimating equation (GEE) models for clustered binary data.     

Identification of biomarkers for early diagnosis of Parkinson's disease by multi-omics joint analysis

ObjectiveThe purpose of this study is to identify the biomarkers for early diagnosis of Parkinson's disease (PD) by multi-omics joint analysis, so as to identify the biomarkers for early diagnosis of PD, and to help clinicians make early diagnosis and treatment.MethodsIn this study, mice are taken as the study subjects. The model of PD mice is established, and then lymphocyte, striatum, substantia nigra protein and proteolysis are extracted. After that, the experiments of protein imprinting and 4¹⁸O labeling are carried out. Mass Spectrometry (MS) analysis technology is mainly used to study proteomics and to analyze the quantitative and qualitative situation of differential proteins in striatum, substantia nigra protein and lymphocyte. By this method, biomarkers for early diagnosis of PD are analyzed and identified.ResultsThe biomarkers of Parkinson's early onset are related to the same quantitative differential expression of lymphocyte, striatum, substantia nigra protein, lymphocyte and substantia nigra.ConclusionThis experimental method can analyze and identify the biomarkers of early diagnosis of PD, help to explore the pathophysiology and pathogenesis of PD, effectively help clinicians make timely diagnosis in advance, and improve the prevention and treatment effect of the disease.     

Estimation of Spatial Variation in Risk Using Matched Case‐control Data

A common problem in environmental epidemiology is to estimate spatial variation in disease risk after accounting for known risk factors. In this paper we consider this problem in the context of matched case‐control studies. We extend the generalised additive model approach of Kelsall and Diggle (1998) to studies in which each case has been individually matched to a set of controls. We discuss a method for fitting this model to data, apply the method to a matched study on perinatal death in the North West Thames region of England and explain why, if spatial variation is of particular scientific interest, matching is undesirable.     

Improved exact confidence intervals for the odds ratio in two independent binomial samples

For two independent binomial samples, the usual exact confidence interval for the odds ratio based on the conditional approach can be very conservative. Recently, Agresti and Min (2002) showed that the unconditional intervals are preferable to conditional intervals with small sample sizes. We use the unconditional approach to obtain a modified interval, which has shorter length, and its coverage probability is closer to and at least the nominal confidence coefficient.     

Selection of biomarkers for HIV-1 latency by integrated analysis

A major obstacle in the treatment of human immunodeficiency virus type 1 (HIV-1) is its ability to establish latent infection. To find novel biomarkers associated with the mechanism of HIV-1 latent infection, we identified 70 candidate genes in HIV-1 latently infected cells through the integrated analysis in a previous study. It is important to select more effective biomarkers among 70 candidates and to verify the possibility of selected biomarkers for HIV-1 latency. We identified the 24 and 25 genes from 70 candidate genes in significantly enriched categories selected by Database for Annotation, Visualization and Integrated Discovery (DAVID) software and Gene Set Enrichment Analysis (GSEA) software, respectively. Also, we investigated genes regulated in both HIV-1 latently infected cell lines and PBMCs from HIV-1 infected patients and found the genes with a common pattern of expression levels in both cell lines and PBMCs. Consequently, we identified nine genes, APBB2, GMPR, IGF2BP3, LRP1, MAD2L2, MX1, OXR1, PTK2B, and TNFSF13B, via integrated analysis. Especially, APBB2 and MAD2L2 were identified in both DAVID and GSEA software. Our findings suggest that nine genes were identified via integrated analysis as potential biomarkers and in particular, APBB2 and MAD2L2 may be considered as more significant biomarkers for HIV-1 latency.