Statistical analysis of the micronucleus test

Various statistical procedures are discussed to compare the incidence of micronucleated polychromatic erythrocytes between various treatment groups and a negative control. The effect of toxicity on the choice of statistical procedure is discussed. The procedure which is recommended is based on likelihood ratio tests comparing each dose group separately against the negative control. The significance level is controlled to allow for the multiple comparisons.     

一个适于最大似然法估计物种分化年代的进化速率平滑近似方法

众所周知 ,物种分化年代的估计对分子钟 (进化速率恒定 )假定很敏感。另一方面 ,在远缘物种 (例如哺乳纲不同目的动物 )的比较中 ,分子钟几乎总是不成立的。这样在估计分化时间时考虑不同进化区系的速率差异至为重要。最大似然法可以很自然地考虑这种速率差异 ,并且可以同时分析多个基因位点的资料以及同时利用多重化石校正数据。以前提出的似然法需要研究者将进化树的树枝按速率分组 ,本文提出一个近似方法以使这个过程自动化。本方法综合了以前的似然法、贝斯法及近似速率平滑法的一些特征。此外 ,还对算法加以改进 ,以适应综合数据分析时某些基因在某些物种中缺乏资料的情形。应用新提出的方法来分析马达加斯加的倭狐猴的分化年代 ,并与以前的似然法及贝斯法的分析进行了比较     

植物QTL定位方法的研究进展

本文系统地介绍了QTL定位的单一标记分析法、区间作图法以及复合区间作图法、混合显性模型的分析方法,概述了一些主要定位方法的分析原理、存在的主要优缺点。单一标记分析法可以采用方差分析、回归分析或似然比检验的方法分析。区间作图法和复合区间作图法是基于两个相邻标记的QTL定位方法,可采用回归分析或最大似然法分析。复合区间作图法在模型中包括了与其他QTL连锁的标记,可以提高作图的精度和效率。混合线性模型的QTL定位方法可以包括复杂的遗传效应及QTL与环境的互作效应,具有更广阔的应用前景。 Abstract:QTL mapping methods are reviewed for single-marker mapping,interval mapping,composite interval mapping,and mixed-model based method.Statistical approaches along with their properties are discussed for the mapping methods.ANOVA,regression method and likelihood ratio test can be applied in single-marker mapping.Interval mapping and composite interval mapping can be conducted,based on two interval markers,by regression method and maximum likelihood method.Since markers linked with other QTLs are include in the model,composite interval mapping is more precision and powerful.Mapping QTL by mixed-model approaches is more applicable when complicated QTL effects as well as QTL by environment interaction are analyzed.     

Detecting simultaneous changepoints in multiple sequences

We discuss the detection of local signals that occur at the same location in multiple one-dimensional noisy sequences, with particular attention to relatively weak signals that may occur in only a fraction of the sequences. We propose simple scan and segmentation algorithms based on the sum of the chi-squared statistics for each individual sample, which is equivalent to the generalized likelihood ratio for a model where the errors in each sample are independent. The simple geometry of the statistic allows us to derive accurate analytic approximations to the significance level of such scans. The formulation of the model is motivated by the biological problem of detecting recurrent DNA copy number variants in multiple samples. We show using replicates and parent-child comparisons that pooling data across samples results in more accurate detection of copy number variants. We also apply the multisample segmentation algorithm to the analysis of a cohort of tumour samples containing complex nested and overlapping copy number aberrations, for which our method gives a sparse and intuitive cross-sample summary.     

分子进化生物学中序列分析方法的新进展

简要介绍了分子进化生物学中序列分析方法的最新进展,特别强调了似然比检验和贝叶斯推论在分子进化和系统发育假说检验中的重要性,并介绍了新方法的一些成功应用,同时还给出了一些重要的信息资源。     

Ovarian structures and uterine environment are associated with phenotypic and genetic merit for performance in lactating dairy cows

The objective of this study was to estimate the association between detailed reproductive phenotypes for cows categorized as divergent for phenotypic and genetic performance. The hypothesis was that higher yielding animals, either phenotypically or genetically, would have compromised ovarian and uterine reproductive performance. Detailed reproductive traits including multiple ovulations, cystic ovarian structures, corpus luteum (CL) presence, and uterine environment were available on 9675 ultrasound records from 8174 dairy lactating cows, calved between 10 and 70 days. Cows were categorized, within parity, into low, average, or high for each of the performance traits. There was a greater likelihood of multiple ovulations in cows with greater phenotypic yields (odds ratio: 1.53–1.81) and greater genetic merit for yield (odds ratio: 1.31–1.59) relative to lower performing contemporaries. After adjustment for genetic merit, a similar trend of increased odds (odds ratio: 1.29–1.87) of multiple ovulations in higher yielding cows was observed compared with the lowest yielding category. There was no association between either phenotypic milk composition or genetic merit for milk composition with the likelihood of multiple ovulations. The likelihood of cystic ovarian structures was highest in cows with greatest phenotypic milk yields (odds ratio: 2.75–3.24), greater genetic merit for milk yield (odds ratio: 1.30–1.51), and even after adjustment for genetic merit there was a greater likelihood of cystic ovarian structures in cows with the highest milk yields (odds ratio: 2.71–2.95), compared with cows in the lowest category for each of the milk traits. Cows with average phenotypic milk yields were more likely to have a CL, compared with the lowest yielding category (odds ratio: 1.20–1.23), and these associations remained after adjustment for genetic merit of the trait. The likelihood of CL presence was highest in cows with the lowest genetic merit for milk. Lower fat:protein ratio was associated with an increased likelihood of CL presence compared with cows with greater fat:protein ratio and cows with the highest phenotypic milk composition were more likely to have a CL compared with cows in the lowest composition category. Genetic predisposition to higher somatic cell score was associated with a reduced risk of multiple ovulations (odds ratio: 0.69; 95% CI: 0.55–0.87) but an increased likelihood of CL presence (odds ratio: 2.66; 95% CI: 2.09–3.37) and poorer uterine health score (odds ratio: 1.36; 95% CI: 1.20–1.55). There was a lower likelihood of multiple ovulations, cystic ovarian structures, and poorer uterine health and an increased likelihood of CL presence in cows with superior genetic merit for reproductive performance and survival.     

Dramatically elevated rate of mitochondrial substitution in lice (Insecta: Phthiraptera)

Few estimates of relative substitution rates, and the underlying mutation rates, exist between mitochondrial and nuclear genes in insects. Previous estimates for insects indicate a 2-9 times faster substitution rate in mitochondrial genes relative to nuclear genes. Here we use novel methods for estimating relative rates of substitution, which incorporate multiple substitutions, and apply these methods to a group of insects (lice, Order: Phthiraptera). First, we use a modification of copath analysis (branch length regression) to construct independent comparisons of rates, consisting of each branch in a phylogenetic tree. The branch length comparisons use maximum likelihood models to correct for multiple substitution. In addition, we estimate codon-specific rates under maximum likelihood for the different genes and compare these values. Estimates of the relative synonymous substitution rates between a mitochondrial (COI) and nuclear (EF-1alpha) gene in lice indicate a relative rate of several 100 to 1. This rapid relative mitochondrial rate (>100 times) is at least an order of magnitude faster than previous estimates for any group of organisms. Comparisons using the same methods for another group of insects (aphids) reveals that this extreme relative rate estimate is not simply attributable to the methods we used, because estimates from aphids are substantially lower. Taxon sampling affects the relative rate estimate, with comparisons involving more closely related taxa resulting in a higher estimate. Relative rate estimates also increase with model complexity, indicating that methods accounting for more multiple substitution estimate higher relative rates.     

Detecting differential expression in microarray data: comparison of optimal procedures

Background  

Many procedures for finding differentially expressed genes in microarray data are based on classical or modified t-statistics. Due to multiple testing considerations, the false discovery rate (FDR) is the key tool for assessing the significance of these test statistics. Two recent papers have generalized two aspects: Storey et al. (2005) have introduced a likelihood ratio test statistic for two-sample situations that has desirable theoretical properties (optimal discovery procedure, ODP), but uses standard FDR assessment; Ploner et al. (2006) have introduced a multivariate local FDR that allows incorporation of standard error information, but uses the standard t-statistic (fdr2d). The relationship and relative performance of these methods in two-sample comparisons is currently unknown.     

Estimating the Transition/Transversion Ratio from Independent Pairwise Comparisons with an Assumed Phylogeny

A method is presented for estimating the transition/transversion ratio (TI/TV), based on phylogenetically independent comparisons. TI/TV is a parameter of some models used in phylogeny estimation intended to reflect the fact that nucleotide substitutions are not all equally likely. Previous attempts to estimate TI/TV have commonly faced three problems: (1) few taxa; (2) nonindependence among pairwise comparisons; and (3) multiple hits make the apparent TI/TV between two sequences decrease over time since their divergence, giving a misleading impression of relative substitution probabilities. We have made use of the time dependency, modeling how the observed TI/TV changes over time and extrapolating to estimate the ``instantaneous' TI/TV—the relevant parameter for phylogenetic inference. To illustrate our method, TI/TV was estimated for two mammalian mitochondrial genes. For 26 pairs of cytochrome b sequences, the estimate of TI/TV was 5.5; 16 pairs of 12s rRNA yielded an estimate of 9.5. These estimates are higher than those given by the maximum likelihood method and than those obtained by averaging all possible pairwise comparisons (with or without a two-parameter correction for multiple substitutions). We discuss strengths, weaknesses, and further uses of our method. Received: 22 August 1995 / Accepted: 26 July 1996     

Preponderance of slightly deleterious polymorphism in mitochondrial DNA: nonsynonymous/synonymous rate ratio is much higher within species than between species

We estimated synonymous (dN) and nonsynonymous (dS) substitution rates for protein-coding genes of the mitochondrial genome from two individuals each of the species human, chimpanzee, and gorilla. The genes were analyzed both separately and in a combined data set. Pairwise sequence comparisons suggest that the dN/dS rate ratios are about 5-10 times higher in within-species comparisons than in between-species comparisons. This result is confirmed by a more rigorous likelihood ratio test, which rejected the null hypothesis that the dN/dS rate ratios are identical within and between species. The likelihood models account for the genetic code structure, transition/transversion rate ratio, and codon usage bias and are expected to produce more reliable results than the commonly used contingency test. Separate analyses of different genes show that the dN/dS rate ratios are higher within species than between species for all 13 mitochondrial genes, with the difference being statistically significant for all except three small or slowly evolving genes. Furthermore, in conserved genes, nonsynonymous rates within species tend to be higher than the between-species rates by a greater proportion than in fast-changing genes. Our findings confirm and extend earlier results obtained from smaller data sets and suggest the operation of slightly deleterious mutations throughout the mitochondrial genome in the hominoids. Implications of the results for evolutionary studies and, in particular, for studies of the origin of modern humans, are discussed.     

Reconstructing phylogenies from allozyme data: comparing method performance with congruence

Allozyme data are widely used to infer the phylogenies of populations and closely-related species. Numerous parsimony, distance, and likelihood methods have been proposed for phylogenetic analysis of these data; the relative merits of these methods have been debated vigorously, but their accuracy has not been well explored. In this study, I compare the performance of 13 phylogenetic methods (six parsimony, six distance, and continuous maximum likelihood) by applying a congruence approach to eight allozyme data sets from the literature. Clades are identified that are supported by multiple data sets other than allozymes (e.g. morphology, DNA sequences), and the ability of different methods to recover these 'known' clades is compared. The results suggest that (1) distance and likelihood methods generally outperform parsimony methods, (2) methods that utilize frequency data tend to perform well, and (3) continuous maximum likelihood is among the most accurate methods, and appears to be robust to violations of its assumptions. These results are in agreement with those from recent simulation studies, and help provide a basis for empirical workers to choose among the many methods available for analysing allozyme characters.     

Relative Risk of Involuntary Injuries Among Currently Employed Adults in the U.S.

It can be very informative to compare the risk of injury for different occupations, places where accidents happen, demographic variables, and other characteristics. However, most epidemiologic literature shows that multiple logistic regression or odds ratio were commonly used to compare the relationship between the variables on two or more levels. A multiple logistic regression could be used for multiple comparisons assuming that the model correctly reflects the situation at hand. But it might not be a true assumption. Assuming no model, we defined the relative risk to compare risks of injury for different occupations, and applied this method to the data collected by the National Center for Health Statistics (NCHS) National Health Interview Survey (NHIS). This application showed that the relative risk of injury varied by occupation and by place of accident, sex, race, and age.     

Unbiased Confidence Intervals for the Odds Ratio of Two Independent Binomial Samples with Application to Case–Control Data

The problem of confidence interval construction for the odds ratio of two independent binomial samples is considered. Two methods of eliminating the nuisance parameter from the exact likelihood, conditioning and maximization, are described. A conditionally exact tail method exists by putting together upper and lower bounds. A shorter interval can be obtained by simultaneous consideration of both tails. We present here new methods that extend the tail and simultaneous approaches to the maximized likelihood. The methods are unbiased and applicable to case-control data, for which the odds ratio is important. The confidence interval procedures are compared unconditionally for small sample sizes in terms of their expected length and coverage probability. A Bayesian confidence interval method and a large-sample chi2 procedure are included in the comparisons.     

Synonymous and nonsynonymous rate variation in nuclear genes of mammals

A maximum likelihood approach was used to estimate the synonymous and nonsynonymous substitution rates in 48 nuclear genes from primates, artiodactyls, and rodents. A codon-substitution model was assumed, which accounts for the genetic code structure, transition/transversion bias, and base frequency biases at codon positions. Likelihood ratio tests were applied to test the constancy of nonsynonymous to synonymous rate ratios among branches (evolutionary lineages). It is found that at 22 of the 48 nuclear loci examined, the nonsynonymous/synonymous rate ratio varies significantly across branches of the tree. The result provides strong evidence against a strictly neutral model of molecular evolution. Our likelihood estimates of synonymous and nonsynonymous rates differ considerably from previous results obtained from approximate pairwise sequence comparisons. The differences between the methods are explored by detailed analyses of data from several genes. Transition/transversion rate bias and codon frequency biases are found to have significant effects on the estimation of synonymous and nonsynonymous rates, and approximate methods do not adequately account for those factors. The likelihood approach is preferable, even for pairwise sequence comparison, because more-realistic models about the mutation and substitution processes can be incorporated in the analysis. Received: 17 May 1997 / Accepted: 28 September 1997     

Some Tests on a Single Change in Independent Copies of Short Sequences: A Cancer Diagnostic Study

Testing a single change in mean or variance when the observations are n independent short sequences has not been fully researched in the literature. Two statistical tests, max t and modified likelihood radtio, are used in this study. Power comparisons and examples are also given. As expected the modified likelihood ratio test yields the largest power. However, with the assumptions we have in the examples, the max t might be the optimal method to use. The max t may prove useful in assessing departures from the steady state in which the data are short sequences of diagnostic information.     

Common Slope Tests for Bivariate Errors‐in‐Variables Models

Likelihood ratio tests are derived for bivariate normal structural relationships in the presence of group structure. These tests may also be applied to less restrictive models where only errors are assumed to be normally distributed. Tests for a common slope amongst those from several datasets are derived for three different cases – when the assumed ratio of error variances is the same across datasets and either known or unknown, and when the standardised major axis model is used. Estimation of the slope in the case where the ratio of error variances is unknown could be considered as a maximum likelihood grouping method. The derivations are accompanied by some small sample simulations, and the tests are applied to data arising from work on seed allometry.     

The Method of Unweighted Means in Unbalanced Analysis of Variance

For factorial experiments with unbalanced data the method of unweighted means is an alternate method of analysis which is computationally simpler than likelihood ratio based F-tests. The quality of the null approximation to the F-distribution of the statistic for the method of unweighted means and the power of this test relative to the likelihood ratio test are discussed. Recommendations are made of when in the course of data analysis the superior operating characteristics of the likelihood ratio test may not outweigh the computational simplicity of the method of unweighted means test.     

Multiple comparisons for analyzing dichotomous response

Dichotomous response models are common in many experimental settings. Statistical parameters of interest are typically the probabilities, pi, that an experimental unit will respond at the various treatment levels. Herein, simultaneous procedures are considered for multiple comparisons among these probabilities, with attention directed at construction of simultaneous confidence intervals for various functions of the pi. The inferences are based on the asymptotic normality of the maximum likelihood estimator of pi. Specific applications include all pairwise comparisons and comparisons with a fixed (control) treatment. Monte Carlo evaluations are undertaken to examine the small-sample properties of the various procedures. It is seen that use of the usual estimates of variance consistently leads to less-than-nominal empirical coverage for most sample sizes examined. For very large samples (total size greater than about 300), nominal coverage is achieved. A reformulation of the pairwise comparisons using a construction noted by Beal (1987, Biometrics 43, 941-950) is shown to exhibit generally nominal empirical coverage characteristics, and is recommended for use with small-to-moderate sample sizes.     

Shift in diversification in sister-clade comparisons: a more powerful test

Tests of shift in diversification associated with key innovations or directional environmental change can be performed with sister-clade comparisons. This approach is attractive because it does not require detailed phylogenetic information. I propose a new likelihood ratio test based on fitting two models of diversification. I show how this test differs from a previous likelihood ratio test based on the geometric distribution. With simulations from a wide range of situations, I show that the new test performs much better than this test and the traditional test by Slowinski and Guyer. The proposed test performs at least as well as the species richness contrast test that has been proposed by several authors in four versions. A power analysis with low number of pairs of sister clades showed that the new test could detect a shift in diversification with five or less pairs of sister clades, whereas the diversity contrast test cannot detect any shift in this situation. The former appears as more powerful than the latter, and therefore is recommended when the number of pairs of sister clades is low (less than 10). All other tests should not be used as the present study showed they lack statistical power and robustness.     

A comparative study of conditional maximum likelihood estimation of a common odds ratio

The finite-sample properties of various point estimators of a common odds ratio from multiple 2 X 2 tables have been considered in a number of simulation studies. However, the conditional maximum likelihood estimator has received only limited attention. That omission is partially rectified here for cases of relatively small numbers of tables and moderate to large within-table sample sizes. The conditional maximum likelihood estimator is found to be superior to the unconditional maximum likelihood estimator, and equal or superior to the Mantel-Haenszel estimator in both bias and precision.