Individual and population penalized regression splines for accelerated longitudinal designs

In an accelerated longitudinal design (ALD), individuals enter the study at different points of their growth trajectory and are observed over a short time span relative to the entire time span of interest. ALD data are combined across independent units to provide an estimate of an overall population curve and predictions of individual patterns of change. As a modest extension of the work of Ruppert et al. (2003, Semiparametric Regression, Cambridge University Press), we develop a computationally efficient procedure for the application of longitudinal semiparametric methods under ALD sampling schemes. We compare balanced and complete longitudinal designs to ALDs using the Berkeley Growth Study data and apply our method to longitudinal magnetic resonance imaging (MRI) brain structure size (volume) measurements from an ongoing developmental study. Potential applications extend beyond growth studies to many other fields in which cost and feasibility constraints impose restrictions on sample size and on the numbers and timings of repeated measurements across subjects.     

Interactions Between Habitat Use and Patterns of Abundance in Coral-dwelling Fishes of the Genus <Emphasis Type="Italic">Gobiodon</Emphasis>

Coral-dwelling fishes from the genus Gobiodon are some of the most habitat specialised fishes on coral reefs. Consequently, we might expect that their population dynamics will be closely associated with the abundance of host corals. I used a combination of log-linear modelling and resource selection ratios to examine patterns of habitat use among eight species of Gobiodon in Kimbe Bay, Papua New Guinea. I then used multiple regression analysis to investigate relationships between the abundance of each species of Gobiodon and the abundance of the corals they inhabited. Each species of Gobiodon used one or more species of coral more frequently than expected by chance. The pattern of habitat use exhibited by each species of Gobiodon did not vary among reef zones or among reefs with different exposures to prevailing winds, despite changes in the relative abundances of corals among reef zones. This consistency in habitat use might be expected if the coral species inhabited confer considerable fitness advantages and, therefore, are strongly preferred. For most species of Gobiodon, abundances among reef zones and exposure regimes were correlated with the abundance of the coral species usually inhabited. Therefore, it appears that habitat availability helps determine abundances of most species of Gobiodon in Kimbe Bay. In addition to correlations with habitat availability, the abundances of G. histrio, G. quinquestrigatus, G. rivulatus (dark form) and the group others were also associated with particular reef zones and exposure regimes. Therefore, in these species, reef type appears to influence patterns of abundance independently of coral availability. In contrast to other species of Gobiodon, the abundance of the most specialised species, Gobiodon sp.A, was not closely associated with the abundance of the only coral species it inhabited. This study demonstrates that even for habitat specialised species, the relationship between habitat availability and abundance varies widely and is multiscale.     

Regression models for convex ROC curves

The performance of a diagnostic test is summarized by its receiver operating characteristic (ROC) curve. Under quite natural assumptions about the latent variable underlying the test, the ROC curve is convex. Empirical data on a test's performance often comes in the form of observed true positive and false positive relative frequencies under varying conditions. This paper describes a family of regression models for analyzing such data. The underlying ROC curves are specified by a quality parameter delta and a shape parameter mu and are guaranteed to be convex provided delta > 1. Both the position along the ROC curve and the quality parameter delta are modeled linearly with covariates at the level of the individual. The shape parameter mu enters the model through the link functions log(p mu) - log(1 - p mu) of a binomial regression and is estimated either by search or from an appropriate constructed variate. One simple application is to the meta-analysis of independent studies of the same diagnostic test, illustrated on some data of Moses, Shapiro, and Littenberg (1993). A second application, to so-called vigilance data, is given, where ROC curves differ across subjects and modeling of the position along the ROC curve is of primary interest.     

Estimation of individual genetic effects from binary observations on relatives applied to a family history of respiratory illnesses and chronic lung disease of newborns

This paper considers methods for estimating the relationship between a binary response Y and the genetic effects responsible for a second binary trait Z. The responses Y are observed only for target individuals, and the responses Z are observed only for the relatives of these targets. The analysis consists of two parts. The first part concerns the analysis of the family data Z and the second part estimates the relation between the genetic effects and Y. For the family data, a generalized linear mixed model with a logit link and Gaussian genetic (random) effects is used. Estimates of the variances of the genetic effects are obtained by using a pseudo-profile log-likelihood method. Estimation of the log likelihood involves averaging over n-dimensional normal distributions, which is done by importance sampling. The methods used in the second part are straightforward. The methods are applied to a data set containing chronic lung disease (CLDN) responses of newborns and asthma (AS), allergy (AL), chronic bronchitis (CB) and eczema (EC) responses observed for the relatives of these newborns. The clinical question is whether genetic effects of AS, AL, CB, and EC have an effect on the risk for CLDN. It can be concluded that for AS, AL, CB, and EC, the influence of genetic effects is significant. However, these genetic predispositions have no significant effect on CLDN.     

A comparison study of models and fitting procedures for biphasic viral dynamics in HIV-1 infected patients treated with antiviral therapies

The study of HIV dynamics is one of the most important developments in recent AIDS research. It has led to a new understanding of the pathogenesis of HIV infection. But, although important findings in HIV dynamics have been published in prestigious scientific journals in the last 5 years, the model-fitting procedures used in these publications have not been studied in any detail. In this paper, we evaluate the performance of four model-fitting procedures proposed and used in biphasic HIV dynamic data analysis via extensive Monte Carlo simulations. We propose some guidelines for practitioners to select an appropriate method for their own data analysis. Real data examples from an AIDS clinical trial are provided as illustrations.     

Measures of spatial heterogeneity for species occurrence or disease incidence with finite-counts

We propose two types of indices with finite-count correction to measure the spatial heterogeneity of binary characteristics of organisms, such as occurrence or non-occurrence of organisms and infected or non-infected plants. We consider the following two examples: plant occurrence in a grassland community, and yellow dwarf disease infection in a rice field. For the grassland community, N quadrats comprising n cells of equal area, were set at random sites in a grassland, and the occurrence of a given species A in each of n cells was recorded. For disease infection, N quadrats, each consisting of n rice plants, were set at random sites in a paddy field, and the number of plants infected with yellow dwarf virus in each quadrat were counted. In these examples, since the number of cells in a quadrat is finite, neither occurrence nor incidence increase infinitely, unlike the number of aphids on a maize leaf. The first category of index belongs to the mean : variance ratio type. The estimated index value for occurrence (or incidence) is the same as that for non-occurrence (or non-incidence). The second category belongs to the k-type of a negative binomial distribution. If some random plants die or recover from the disease then the expected value of the second type of index does not change. For n, the current indices approach the mean : variance ratio and the inverse of k in a negative binomial distribution, respectively. This indicates that these indices are suitable for binary data sets.     

Skeletal estimates of Gigantopithecus based on a Gorilla analogy

Mandibular metrics of Gorilla gorilla are used to generate multiple regression equations using long bones metrics as criterion variables. These equations are then used to predict skeletal estimates of Gigantopithecus based upon mandibular metrics. On the average, the long bones of Gigantopithecus were 20–25% longer and more robust than those of living gorillas.     

Grain growth and endosperm cell size under high night temperatures in rice (Oryza sativa L.)

BACKGROUND AND AIMS: High night temperatures are more harmful to grain weight in rice than high day temperatures. Grain growth rate and growth duration were investigated to determine which was the cause of the decrease in final grain weight under high night temperatures. Endosperm cell number and cell sizes were also examined to determine which might cause the decrease in final grain weight. METHODS: Rice plants were grown outdoors in plastic pots and moved at heading time to three temperature-controlled glasshouses under high night temperature (HNT; 22/34 degrees C), high day temperature (HDT; 34/22 degrees C) and control conditions (CONT; 22/22 degrees C). Grains were sampled periodically, and the time-course of grain growth was divided into rate and duration by logistic regression analysis. Endosperm cell numbers and cell sizes were analysed by digitalized hand-tracing images of endosperm cross-sections. KEY RESULTS: The duration of grain growth was reduced by high temperature both day and night. However, the rate of grain growth was lower in HNT than in HDT. The number of cells in endosperm cross-sections in HNT was similar to that in HDT, and higher than that in CONT. The average cell area was smaller in HNT than in either CONT or HDT. The differences in average cell areas between HNT and HDT were greater at distances 60-80 % from the central point of endosperm towards the endosperm surface. CONCLUSIONS: The results show that HNT compared with HDT reduced the final grain weight by a reduction in grain growth rate in the early or middle stages of grain filling, and also reduced cell size midway between the central point and the surface of endosperm.