Statistical methods for detecting molecular adaptation

The past few years have seen the development of powerful statistical methods for detecting adaptive molecular evolution. These methods compare synonymous and nonsynonymous substitution rates in protein-coding genes, and regard a nonsynonymous rate elevated above the synonymous rate as evidence for darwinian selection. Numerous cases of molecular adaptation are being identified in various systems from viruses to humans. Although previous analyses averaging rates over sites and time have little power, recent methods designed to detect positive selection at individual sites and lineages have been successful. Here, we summarize recent statistical methods for detecting molecular adaptation, and discuss their limitations and possible improvements.     

Electrogenic Properties of the Sodium Pump in a Dynamic Model of Membrane Transport

The general purpose of this theoretical work is to contribute to understand the physiological role of the electrogenic properties of the sodium pump, by studying a dynamic model that integrates diverse processes of ionic and water transport across the plasma membrane. For this purpose, we employ a mathematical model that describes the rate of change of the intracellular concentrations of Na⁺, K⁺ and Cl⁻, of the cell volume, and of the plasma membrane potential (V _m ). We consider the case of a nonexcitable, nonpolarized cell expressing the sodium pump; Na⁺, K⁺, Cl⁻ and water channels, and cotransporters of KCl and NaCl in its plasma membrane. We particularly analyze here the conditions under which the physiological V _m can be generated in a predominantly electrogenic fashion, as a result of the activity of the sodium pump. A major conclusion of this study is that, for the cell model considered, a low potassium permeability is not a sufficient condition for a predominantly electrogenic generation of the V _m by the sodium pump. The presence of an electroneutral exchange of Na⁺ and K⁺ represents a necessary additional requirement. Received: 8 September 1999/Revised: 21 March 2000     

Small-sample bias and corrections for conditional maximum-likelihood odds-ratio estimators

A number of small-sample corrections have been proposed for the conditional maximum-likelihood estimator of the odds ratio for matched pairs with a dichotomous exposure. I here contrast the rationale and performance of several corrections, specifically those that generalize easily to multiple conditional logistic regression. These corrections or Bayesian analyses with informative priors may serve as diagnostics for small-sample problems. Points are illustrated with a small exact performance comparison and with an example from a study of electrical wiring and childhood leukemia. The former comparison suggests that small-sample bias may be more prevalent than commonly realized.     

Lattice Model for QSAR Studies

A system of lattice models that takes into account the structures of molecules, their form, stereochemical features and their interaction with the enclosing space, is proposed. The local, integral and field structural parameters of molecules (more than 20 thousand per compound) are estimated within the proposed framework. An investigation of the utility of these parameters in Quantitative Structure-Activity Relationships (QSAR) has been made using several statistical methods (multiple regression analysis, partial least squares (PLS), trend-vector procedure). The efficiency of the proposed approach has been examined using a data set derived from the formation of charge-transfer complexes of monosubstituted bezens with 1,3,5-trinitrobenzene.     

Mitochondrial DNA (mtDNA) reveals that female Bechstein's bats live in closed societies

We present a microgeographic analysis of mitochondrial DNA (mtDNA) in Bechstein's bats using three sources of control region sequence variability, including a novel mtDNA microsatellite, to assess individual relatedness both within and among 10 maternity colonies. Comparison of marker variability among 268 adult females revealed little genetic variability within each colony. However, most colonies were clearly distinguished by colony-specific mitochondrial haplotypes (total n = 28). Low intracolony variability and strong haplotype segregation among colonies, was reflected by an extraordinary high FST of 0.68, indicating a very low intercolony dispersal rate of approximately one female in five generations. Haplotype distribution among 18 solitary males showed that males frequently disperse between colony locations, indicating the absence of dispersal barriers. Bechstein's bat maternity colonies are thus closed groups that comprise 20-40 females probably belonging to only one or, at most, two matrilines. The genetic population structure of Bechstein's bats is in agreement with the hypothesis that females seek familiar and, at least, partially related cooperation partners for raising their young. Alternatively strong philopatry might reflect the importance of profound roost or habitat knowledge for successful reproduction in female Bechstein's bats.     

Multiple origins of eusociality among sponge-dwelling shrimps (Synalpheus)

Abstract.— As the most extreme expression of apparent altruism in nature, eusociality has long posed a central paradox for behavioral and evolutionary ecology. Because eusociality has arisen rarely among animals, understanding the selective pressures important in early stages of its evolution remains elusive. Employing a historical approach to this problem, we used morphology and DNA sequences to reconstruct the phylogeny of 13 species of sponge-dwelling shrimps ( Synalpheus ) with colony organization ranging from asocial pair-bonding through eusociality. We then used phylogenetically independent contrasts to test whether sociality was associated with evidence of enhanced competitive ability, as suggested by hypotheses invoking an advantage of cooperation in crowded habitats. The molecular, morphological, and combined data each strongly supported three independent origins of monogynous, multigenerational (eusocial) colony organization within this genus. Phylogenetically independent contrasts confirmed that highly social taxa, with strong reproductive skew, have significantly higher relative abundance within the host sponge than do less social taxa, a result that was robust to uncertainty in tree topology and varying models of character change. A similar tendency for highly social species to share their sponge with fewer congener species was suggestive, but not significant. Because unoccupied habitat appears to be limiting for many sponge-dwelling shrimp species, these data are consistent with hypotheses that cooperative social groups enjoy a competitive advantage over less organized groups or individuals, where independent establishment is difficult, and that enemy pressure is of central importance in the evolution of animal sociality.     

A diagnostic for Cox regression with discrete failure-time models

Changes in maximum likelihood parameter estimates due to deletion of individual observations are useful statistics, both for regression diagnostics and for computing robust estimates of covariance. For many likelihoods, including those in the exponential family, these delete-one statistics can be approximated analytically from a one-step Newton-Raphson iteration on the full maximum likelihood solution. But for general conditional likelihoods and the related Cox partial likelihood, the one-step method does not reduce to an analytic solution. For these likelihoods, an alternative analytic approximation that relies on an appropriately augmented design matrix has been proposed. In this paper, we extend the augmentation approach to explicitly deal with discrete failure-time models. In these models, an individual subject may contribute information at several time points, thereby appearing in multiple risk sets before eventually experiencing a failure or being censored. Our extension also allows the covariates to be time dependent. The new augmentation requires no additional computational resources while improving results.     

On power and sample size calculations for likelihood ratio tests in generalized linear models

A direct extension of the approach described in Self, Mauritsen, and Ohara (1992, Biometrics 48, 31-39) for power and sample size calculations in generalized linear models is presented. The major feature of the proposed approach is that the modification accommodates both a finite and an infinite number of covariate configurations. Furthermore, for the approximation of the noncentrality of the noncentral chi-square distribution for the likelihood ratio statistic, a simplification is provided that not only reduces substantial computation but also maintains the accuracy. Simulation studies are conducted to assess the accuracy for various model configurations and covariate distributions.     

Calibration for nonlinear mixed effects models: an application to the withdrawal time prediction

We propose calibration methods for nonlinear mixed effects models. Using an estimator whose asymptotic properties are known, four different statistics are used to perform the calibration. Simulations are carried out to compare the performance of these statistics. Finally, the milk discard time prediction of an antibiotic, which has motivated this study, is performed on real data.