A special distribution is suggested for the analysis of survival data in which there is a long random delay before the onset of the terminal process. Estimation by the method of moments and by maximum likelihood is compared.

service Email alerting click here top right corner of the article or Receive free email alerts when new articles cite this article-sign up in the box at the

Oxidative damage to DNA has been associated with neurodegenerative diseases. Developmental exposure to lead (Pb) has been shown to elevate the Alzheimer's disease (AD) related beta-amyloid peptide (Abeta), which is known to generate reactive oxygen species in the aging brain. This study measures the lifetime cerebral 8-hydroxy-2'-deoxyguanosine (oxo8dG) levels and the activity of the DNA repair enzyme 8-oxoguanine DNA glycosylase (Ogg1) in rats developmentally exposed to Pb. Oxo8dG was transiently modulated early in life (Postnatal day 5), but was later elevated 20 months after exposure to Pb had ceased, while Ogg1 activity was not altered. Furthermore, an age-dependent loss in the inverse correlation between Ogg1 activity and oxo8dG accumulation was observed. The effect of Pb on oxo8dG levels did not occur if animals were exposed to Pb in old age. These increases in DNA damage occurred in the absence of any Pb-induced changes in copper/zinc-superoxide dismutase (SOD1), manganese-SOD (SOD2), and reduced-form glutathion (GSH). These data suggest that oxidative damage and neurodegeneration in the aging brain could be impacted by the developmental disturbances.

SummaryThe case series model for estimating the association between an age-dependent exposure and an outcome event requires information only on cases and implicitly adjusts for all age-independent multiplicative confounders, while allowing for an age-dependent base-line incidence. In the paper the model is presented in greater generality than hitherto, including more general discussion of its derivation, underlying assumptions, applicability, limitations and efficiency. A semiparametric version of the model is developed, in which the age-specific relative incidence is left unspecified. Modelling covariate effects and testing assumptions are discussed. The small sample performance of this model is studied in simulations. The methods are illustrated with several examples from epidemiology.

Microarrays are new biotechnological devices that permit the simultaneous evaluation of expression levels of thousands of genes in one or more tissue samples. We develop a new method for identifying differentially expressed genes in replicated cDNA and oligonucleotide microarray experiments. The method is based on a nonparametric prediction interval which is computed as an order statistic of n control measurements and is applied sequentially to a series of p replicate sets of experimental measurements, each of size ni. We illustrate how reasonable experiment-wise false positive and false negative rates can be attained for any practical number of genes based on manipulating the order statistics, n, p and ni. The method is used to identify gene expression levels that are associated with a pathological condition beyond chance expectations given the large number of genes tested. We illustrate use of the method on replicated gene expression data in tumor and normal colon tissues, and compare it to an alternative approach based on permutation tests.

The mapping and sequencing of the human genome promises rapid growth in understanding the genetically influenced mechanisms that underlie human disease. To realize this promise fully, it is necessary to relate genetic information to clinical phenotypes. Genetic tissue banking in clinical studies provides opportunities to analyze the genetic contribution to variation in response to treatments. The challenges to progress are likely to come from the complex organizational, social, political, and ethical issues that must be resolved in order to put clinical and DNA bank information together. Concerns about subjects' rights, informed consent, privacy, and ownership of genetic material require attention in the development of DNA banks. In this paper we describe one approach to the solution of these problems that was adopted by one clinical trials group, the Department of Veterans Affairs Cooperative Studies Program.

A high resolution map of the human genome previously has been constructed by using the G3 panel of human/hamster radiation hybrid cell lines and >15,000 unique human genetic markers. By determining whether human DNA sequences are present or absent in each of the hybrids, localization of single genes may routinely be achieved at approximately 250-kb resolution. In this paper we have tested whether similarly precise localization might be achieved by phenotypic screening of the hybrids to facilitate positional cloning of unknown genes. We assayed the susceptibility of each of the hybrid cell lines to transduction by retroviral vectors bearing different retroviral envelope proteins that recognize receptors present on human but not on hamster cells. The results for each of the retroviral vectors were informative and allowed precise localization of the receptor genes for the RD114 cat endogenous retrovirus, xenotropic murine leukemia virus, and type C feline leukemia virus. After cloning of the receptors for these retroviruses, we found that standard genotypic mapping by PCR gave results that were nearly identical to those from phenotypic mapping. These experiments show that precise gene localization by phenotypic assay of radiation hybrids is practical and was not appreciably impacted by the known instability of such hybrid cells. This technique should be applicable to many other human genes having discernible phenotypes in hamster cells and, with completion of the human genome project, will allow rapid identification of unknown genes on the basis of phenotype.

We consider likelihood-based asymptotic inference for a p-dimensional parameter theta of an identifiable parametric model with singular information matrix of rank p - 1 at theta=theta* and likelihood differentiable up to a specific order. We derive the asymptotic distribution of the likelihood ratio test statistics for the simple null hypothesis that theta = theta* and of the maximum likelihood estimator (MLE) of theta when theta = theta*. We show that there exists a reparametrization such that the MLE of the last p - 1 components of theta converges at rate O-p(n(-1/2)). For the first component theta(1) of theta the rate of convergence depends on the order s of the first non-zero partial derivative of the log-likelihood with respect to theta(1) evaluated at theta*, When s is odd the rate of convergence of the MLE of theta(1) is O-p(n(-1/2s)). When s is even, the rate of convergence of the MLE of \theta(1)- theta(1)*\ is O-p(n(-1/2s)) and moreover, the asymptotic distribution of the sign of the MLE of theta(1) - theta(1)* is non-standard. When p = 1 it is determined by the sign of the sum of the residuals from the population least-squares regression of the (s + l)th derivative of the individual contributions to the log-likelihood on their derivatives of order s. For p>1, it is determined by a linear combination of the sum of residuals of a multivariate population least-squares regression involving partial and mixed derivatives of the log-likelihood of a specific order. Thus although the MLE of \theta(1) - theta(1)*\ has a uniform rate of convergence of O-p(n(-1/2s)), the uniform convergence rate for the MLE of theta(1) in suitable shrinking neighbourhoods of theta(1)* is only O-p(n(-1/(2s+2))).