Developments from analysis of variance through to generalized linear models and beyond

The collaborations between statisticians and biologists during the 100 years since AAB was founded have led to a very impressive list of statistical techniques, whose use now goes well beyond agriculture and biology. One example is the maximum likelihood methodology for probit analysis, arising from the collaboration between Sir Ronald Fisher and Chester Bliss. Others include analysis of variance, design of experiments, generalized linear models and the residual, or restricted, maximum likelihood (REML) algorithm for fitting unbalanced linear mixed models.     

Sensitivity analysis of transient population dynamics

Short-term, transient population dynamics can differ in important ways from long-term asymptotic dynamics. Just as perturbation analysis (sensitivity and elasticity) of the asymptotic growth rate reveals the effects of the vital rates on long-term growth, the perturbation analysis of transient dynamics can reveal the determinants of short-term patterns. In this article, I present a completely new approach to transient sensitivity and elasticity analysis, using methods from matrix calculus. Unlike previous methods, this approach applies not only to linear time-invariant models but also to time-varying, subsidized, stochastic, nonlinear and spatial models. It is computationally simple, and does not require calculation of eigenvalues or eigenvectors. The method is presented along with applications to plant and animal populations.     

Methods for conducting sensitivity analysis of trials with potentially nonignorable competing causes of censoring

We consider inference for the treatment-arm mean difference of an outcome that would have been measured at the end of a randomized follow-up study if, during the course of the study, patients had not initiated a nonrandomized therapy or dropped out. We argue that the treatment-arm mean difference is not identified unless unverifiable assumptions are made. We describe identifying assumptions that are tantamount to postulating relationships between the components of a pattern-mixture model but that can also be interpreted as imposing restrictions on the cause-specific censoring probabilities of a selection model. We then argue that, although sufficient for identification, these assumptions are insufficient for inference due to the curse of dimensionality. We propose reducing dimensionality by specifying semiparametric cause-specific selection models. These models are useful for conducting a sensitivity analysis to examine how inference for the treatment-arm mean difference changes as one varies the magnitude of the cause-specific selection bias over a plausible range. We provide methodology for conducting such sensitivity analysis and illustrate our methods with an analysis of data from the AIDS Clinical Trial Group (ACTG) study 002.     

A novel statistical analysis and interpretation of flow cytometry data

A recently developed class of models incorporating the cyton model of population generation structure into a conservation-based model of intracellular label dynamics is reviewed. Statistical aspects of the data collection process are quantified and incorporated into a parameter estimation scheme. This scheme is then applied to experimental data for PHA-stimulated CD4+T and CD8+T cells collected from two healthy donors. This novel mathematical and statistical framework is shown to form the basis for accurate, meaningful analysis of cellular behaviour for a population of cells labelled with the dye carboxyfluorescein succinimidyl ester and stimulated to divide.     

Molecular dynamics and free energy analysis of neuraminidase-ligand interactions

We report molecular dynamics calculations of neuraminidase in complex with an inhibitor, 4-amino-2-deoxy-2,3-didehydro-N-acetylneuraminic acid (N-DANA), with subsequent free energy analysis of binding by using a combined molecular mechanics/continuum solvent model approach. A dynamical model of the complex containing an ionized Glu119 amino acid residue is found to be consistent with experimental data. Computational analysis indicates a major van der Waals component to the inhibitor-neuraminidase binding free energy. Based on the N-DANA/neuraminidase molecular dynamics trajectory, a perturbation methodology was used to predict the binding affinity of related neuraminidase inhibitors by using a force field/Poisson-Boltzmann potential. This approach, incorporating conformational search/local minimization schemes with distance-dependent dielectric or generalized Born solvent models, correctly identifies the most potent neuraminidase inhibitor. Mutation of the key ligand four-substituent to a hydrogen atom indicates no favorable binding free energy contribution of a hydroxyl group; conversely, cationic substituents form favorable electrostatic interactions with neuraminidase. Prospects for further development of the method as an analysis and rational design tool are discussed.     

Overdispersed fungus germination data: statistical analysis using R

ABSTRACT

Proportion data from dose-response experiments are often overdispersed, characterised by a larger variance than assumed by the standard binomial model. Here, we present different models proposed in the literature that incorporate overdispersion. We also discuss how to select the best model to describe the data and present, using R software, specific code used to fit and interpret binomial, quasi-binomial, beta-binomial, and binomial-normal models, as well as to assess goodness-of-fit. We illustrate applications of these generalized linear models and generalized linear mixed models with a case study from a biological control experiment, where different isolates of Isaria fumosorosea (Hypocreales: Cordycipitaceae) were used to assess which ones presented higher resistance to UV-B radiation. We show how to test for differences between isolates and also how to statistically group isolates presenting a similar behaviour.     

Feature-specific penalized latent class analysis for genomic data

Genomic data are often characterized by a moderate to large number of categorical variables observed for relatively few subjects. Some of the variables may be missing or noninformative. An example of such data is loss of heterozygosity (LOH), a dichotomous variable, observed on a moderate number of genetic markers. We first consider a latent class model where, conditional on unobserved membership in one of k classes, the variables are independent with probabilities determined by a regression model of low dimension q. Using a family of penalties including the ridge and LASSO, we extend this model to address higher-dimensional problems. Finally, we present an orthogonal map that transforms marker space to a space of "features" for which the constrained model has better predictive power. We demonstrate these methods on LOH data collected at 19 markers from 93 brain tumor patients. For this data set, the existing unpenalized latent class methodology does not produce estimates. Additionally, we show that posterior classes obtained from this method are associated with survival for these patients.     

Fixed point sensitivity analysis of interacting structured populations

Sensitivity analysis of structured populations is a useful tool in population ecology. Historically, methodological development of sensitivity analysis has focused on the sensitivity of eigenvalues in linear matrix models, and on single populations. More recently there have been extensions to the sensitivity of nonlinear models, and to communities of interacting populations. Here we derive a fully general mathematical expression for the sensitivity of equilibrium abundances in communities of interacting structured populations. Our method yields the response of an arbitrary function of the stage class abundances to perturbations of any model parameters. As a demonstration, we apply this sensitivity analysis to a two-species model of ontogenetic niche shift where each species has two stage classes, juveniles and adults. In the context of this model, we demonstrate that our theory is quite robust to violating two of its technical assumptions: the assumption that the community is at a point equilibrium and the assumption of infinitesimally small parameter perturbations. Our results on the sensitivity of a community are also interpreted in a niche theoretical context: we determine how the niche of a structured population is composed of the niches of the individual states, and how the sensitivity of the community depends on niche segregation.     

Developments in statistical analysis in quantitative genetics

A remarkable research impetus has taken place in statistical genetics since the last World Conference. This has been stimulated by breakthroughs in molecular genetics, automated data-recording devices and computer-intensive statistical methods. The latter were revolutionized by the bootstrap and by Markov chain Monte Carlo (McMC). In this overview a number of specific areas are chosen to illustrate the enormous flexibility that McMC has provided for fitting models and exploring features of data that were previously inaccessible. The selected areas are inferences of the trajectories over time of genetic means and variances, models for the analysis of categorical and count data, the statistical genetics of a model postulating that environmental variance is partly under genetic control, and a short discussion of models that incorporate massive genetic marker information. We provide an overview of the application of McMC to study model fit, and finally, a discussion is presented on the development of efficient McMC updating schemes for non-standard models.     

Generation interval contraction and epidemic data analysis

The generation interval is the time between the infection time of an infected person and the infection time of his or her infector. Probability density functions for generation intervals have been an important input for epidemic models and epidemic data analysis. In this paper, we specify a general stochastic SIR epidemic model and prove that the mean generation interval decreases when susceptible persons are at risk of infectious contact from multiple sources. The intuition behind this is that when a susceptible person has multiple potential infectors, there is a "race" to infect him or her in which only the first infectious contact leads to infection. In an epidemic, the mean generation interval contracts as the prevalence of infection increases. We call this global competition among potential infectors. When there is rapid transmission within clusters of contacts, generation interval contraction can be caused by a high local prevalence of infection even when the global prevalence is low. We call this local competition among potential infectors. Using simulations, we illustrate both types of competition. Finally, we show that hazards of infectious contact can be used instead of generation intervals to estimate the time course of the effective reproductive number in an epidemic. This approach leads naturally to partial likelihoods for epidemic data that are very similar to those that arise in survival analysis, opening a promising avenue of methodological research in infectious disease epidemiology.     

Hierarchical multivariate mixture generalized linear models for the analysis of spatial data: An application to disease mapping

Disease mapping of a single disease has been widely studied in the public health setup. Simultaneous modeling of related diseases can also be a valuable tool both from the epidemiological and from the statistical point of view. In particular, when we have several measurements recorded at each spatial location, we need to consider multivariate models in order to handle the dependence among the multivariate components as well as the spatial dependence between locations. It is then customary to use multivariate spatial models assuming the same distribution through the entire population density. However, in many circumstances, it is a very strong assumption to have the same distribution for all the areas of population density. To overcome this issue, we propose a hierarchical multivariate mixture generalized linear model to simultaneously analyze spatial Normal and non‐Normal outcomes. As an application of our proposed approach, esophageal and lung cancer deaths in Minnesota are used to show the outperformance of assuming different distributions for different counties of Minnesota rather than assuming a single distribution for the population density. Performance of the proposed approach is also evaluated through a simulation study.     

Multivariate meta‐analysis with an increasing number of parameters

Meta‐analysis can average estimates of multiple parameters, such as a treatment's effect on multiple outcomes, across studies. Univariate meta‐analysis (UVMA) considers each parameter individually, while multivariate meta‐analysis (MVMA) considers the parameters jointly and accounts for the correlation between their estimates. The performance of MVMA and UVMA has been extensively compared in scenarios with two parameters. Our objective is to compare the performance of MVMA and UVMA as the number of parameters, p, increases. Specifically, we show that (i) for fixed‐effect (FE) meta‐analysis, the benefit from using MVMA can substantially increase as p increases; (ii) for random effects (RE) meta‐analysis, the benefit from MVMA can increase as p increases, but the potential improvement is modest in the presence of high between‐study variability and the actual improvement is further reduced by the need to estimate an increasingly large between study covariance matrix; and (iii) when there is little to no between‐study variability, the loss of efficiency due to choosing RE MVMA over FE MVMA increases as p increases. We demonstrate these three features through theory, simulation, and a meta‐analysis of risk factors for non‐Hodgkin lymphoma.     

A physical analysis of the ion parametric resonance model

We show, in elementary terms, using for the most part only elementary mathematics, the physical bases for the ion parametric resonance model so as to clarify the assumptions and consequences of the model. The analysis shows why, contrary to earlier conclusions, no combination of weak DC and AC magnetic fields can modify the transition rate to the ground state of excited ions. Although reinterpretations of the biological consequences of the motion of the excited ions circumvent that particular objection to the model, those changes introduce other difficulties. Also, other objections to the mechanism still stand; hence the model cannot account for any purported biological effects of weak extremely low frequency magnetic fields. Bioelectromagnetics 19:181–191, 1998. © 1998 Wiley-Liss, Inc.     

Time-varying coefficient models for the analysis of air pollution and health outcome data

In this article a time-varying coefficient model is developed to examine the relationship between adverse health and short-term (acute) exposure to air pollution. This model allows the relative risk to evolve over time, which may be due to an interaction with temperature, or from a change in the composition of pollutants, such as particulate matter, over time. The model produces a smooth estimate of these time-varying effects, which are not constrained to follow a fixed parametric form set by the investigator. Instead, the shape is estimated from the data using penalized natural cubic splines. Poisson regression models, using both quasi-likelihood and Bayesian techniques, are developed, with estimation performed using an iteratively re-weighted least squares procedure and Markov chain Monte Carlo simulation, respectively. The efficacy of the methods to estimate different types of time-varying effects are assessed via a simulation study, and the models are then applied to data from four cities that were part of the National Morbidity, Mortality, and Air Pollution Study.     

An analysis of neutral-alleles and variable-environment diffusion models

Three diffusion models are formulated for the evolution of a diploid population with K alleles at one locus with completely symmetric mutation and random genetic drift, a variable-environment, and all the above mechanisms. For the diallelic case, the transient behavior is studied by solving the corresponding diffusion equations by an asymptotic method valid for short time intervals. The transient behavior of the three models is compared for the case when their stationary distributions are identical. The expected amount of heterozygosity is computed using the asymptotic solution and is compared to an exact result. The asymptotic results are extended to the general case with K alleles at the locus for the symmetric mutation and variable-environment models.Research supported by the National Science Foundation under Grant MCS 79-01718     

A Bayesian analysis of multivariate doubly-interval-censored dental data

A Bayesian survival analysis is presented to examine the effect of fluoride-intake on the time to caries development of the permanent first molars in children between 7 and 12 years of age using a longitudinal study conducted in Flanders. Three problems needed to be addressed. Firstly, since the emergence time of a tooth and the time it experiences caries were recorded yearly, the time to caries is doubly interval censored. Secondly, due to the setup of the study, many emergence times were left-censored. Thirdly, events on teeth of the same child are dependent. Our Bayesian analysis is a modified version of the intensity model of Harkanen et al. (2000, Scandinavian Journal of Statistics 27, 577-588). To tackle the problem of the large number of left-censored observations a similar Finnish data set was introduced. Our analysis shows no convincing effect of fluoride-intake on caries development.     

Evolutionary factor analysis of replicated time series

Summary In this article, we develop a novel method that explains the dynamic structure of multi-channel electroencephalograms (EEGs) recorded from several trials in a motor-visual task experiment. Preliminary analyses of our data suggest two statistical challenges. First, the variance at each channel and cross-covariance between each pair of channels evolve over time. Moreover, the cross-covariance profiles display a common structure across all pairs, and these features consistently appear across all trials. In the light of these features, we develop a novel evolutionary factor model (EFM) for multi-channel EEG data that systematically integrates information across replicated trials and allows for smoothly time-varying factor loadings. The individual EEGs series share common features across trials, thus, suggesting the need to pool information across trials, which motivates the use of the EFM for replicated time series. We explain the common co-movements of EEG signals through the existence of a small number of common factors. These latent factors are primarily responsible for processing the visual-motor task which, through the loadings, drive the behavior of the signals observed at different channels. The estimation of the time-varying loadings is based on the spectral decomposition of the estimated time-varying covariance matrix.     

Chiral symmetry breaking: Experimental results and computer analysis of a liquid-phase autoxidation

The autoxidation of tetralin is treated as a model reaction system to define the applicability of stereospecific autocatalysis. This concept, predicting a spontaneous amplification of enantiomeric excess generated by an autocatalytic chemical reaction, is used in several theoretical models as an explanation for the origin of natural optical activity. The reaction system investigated obeys the basic criteria of these models: a chiral intermediate (tetralin hydroperoxide) is produced from an achiral substrate (tetralin) via an autocatalytic pathway where the feedback mechanism is expected to generate a state of broken chiral symmetry. In order to test the amplification capacity of this reaction a computer analysis of the kinetic scheme is performed. This simulation is derived from the known kinetic scheme of autoxidation and is validated by fitting the experimentally observed data of hydroperoxide evolution. Calculations show that this model allows powerful amplification of enantiomeric excess and a transient amplification of the optical rotation. It is also demonstrated that the model system exhibits pronounced sensitivity toward any loss of absolute configuration of the involved chiral species. Since an amplification effect results exclusively at a high degree of stereoselectivity, it is concluded that stereospecific autocatalysis is possible in systems which show template reactions, crystallization, or colloidal effects. © 1993 Wiley-Liss, Inc.