Accuracy and Power of the Likelihood Ratio Test for Comparing Evolutionary Rates Among Genes

Sequences for multiple protein-coding genes are now commonly available from several, often closely related species. These data sets offer intriguing opportunities to test hypotheses regarding whether different types of genes evolve under different selective pressures. Although maximum likelihood (ML) models of codon substitution that are suitable for such analyses have been developed, little is known about the statistical properties of these tests. We use a previously developed fixed-sites model and computer simulations to examine the accuracy and power of the likelihood ratio test (LRT) in comparing the nonsynonymous-to-synonymous substitution rate ratio (=dN/dS) between two genes. Our results show that the LRT applied to fixed-sites models may be inaccurate in some cases when setting significance thresholds using a ² approximation. Instead, we use a parametric bootstrap to describe the distribution of the LRT statistic for fixed-sites models and examine the power of the test as a function of sampling variables and properties of the genes under study. We find that the power of the test is high (>80%) even when sampling few taxa (e.g., six species) if sequences are sufficiently diverged and the test is largely unaffected by the tree topology used to simulate data. Our simulations show fixed-sites models are suitable for comparing substitution parameters among genes evolving under even strong evolutionary constraint ( 0.05), although relative rate differences of 25% or less may be difficult to detect.Reviewing Editor: Dr. Rosmus Nielsen     

The Residual Likelihood Ratio Test for the Variance Ratio in a Linear Model with Two Variance Components

Residual maximum likelihood has proved to be a successful approach to the estimation of variance components. In this paper, its counterpart in testing, the residual likelihood ratio test, is applied to testing the ratio of two variance components. The test is compared with the Wald test and the locally most powerful invariant test.     

Bootstrap Restricted Likelihood Ratio Test for the Detection of Rare Variants

In this paper the detection of rare variants association with continuous phenotypes of interest is investigated via the likelihood-ratio based variance component test under the framework of linear mixed models. The hypothesis testing is challenging and nonstandard, since under the null the variance component is located on the boundary of its parameter space. In this situation the usual asymptotic chisquare distribution of the likelihood ratio statistic does not necessarily hold. To circumvent the derivation of the null distribution we resort to the bootstrap method due to its generic applicability and being easy to implement. Both parametric and nonparametric bootstrap likelihood ratio tests are studied. Numerical studies are implemented to evaluate the performance of the proposed bootstrap likelihood ratio test and compare to some existing methods for the identification of rare variants. To reduce the computational time of the bootstrap likelihood ratio test we propose an effective approximation mixture for the bootstrap null distribution. The GAW17 data is used to illustrate the proposed test.     

On Computing a Likelihood Ratio Test Statistic in Physiological Response Slope Studies

We consider a structural model often used as the basis for analysis of response slopes in physiological studies, with particular emphasis on the problem of comparing the slopes for two samples. We simplify some of the calculations leading to a likelihood ratio test statistic for the hypothesis of equality of slopes. We show that these simplifications introduce additional computational stability, and illustrate our remarks in a numerical example.     

Normal Approximation of a Discrete Distribution on Exact Test of Uniform Association in a 2 × 3 Table

For the analysis of 2 × 3 tables, TOMIZAWA (1993) considered an exact test of uniform association, which is an extension of independence, and then derived a discrete distribution. This paper gives a normal approximation of the discrete distribution and describes that the normalized statistic can test a one-sided hypothesis on the uniform association. Also it points out that the square of the normalized test statistic is equal to the Pearson's chi-squared statistic for testing the uniform association.     

A Note on Maximum Likelihood Ratio Test of No Outliers in Regression Models

A derivation of the maximum likelihood ratio test for testing no outliers in regression models is given using the method of WETHERILL (1981, pp. 106–107) for estimating the regression parameters. This method is essentially similar to the one outlined in BARNETT and LEWIS (1978, p. 263), although by our detailed derivation it is easier to see that the maximum likelihood estimate of θ of model (3) under the hypothesis that the i^th observation in an outlier is the same as that obtained from model (1) when the i^th observation is removed.     

A Likelihood Ratio Test for the Nonparametric Behrens‐Fisher Problem

The nonparametric Behrens‐Fisher hypothesis is the most appropriate null hypothesis for the two‐sample comparison when one does not wish to make restrictive assumptions about possible distributions. In this paper, a numerical approach is described by which the likelihood ratio test can be calculated for the nonparametric Behrens‐Fisher problem. The approach taken here effectively reduces the number of parameters in the score equations to one by using a recursive formula for the remaining parameters. The resulting single dimensional problem can be solved numerically. The power of the likelihood ratio test is compared by simulation to that of a generalized Wilcoxon test of Brunner and Munzel. The tests have similar power for all alternatives considered when a simulated null distribution is used to generate cutoff values for the tests. The methods are illustrated on data on shoulder pain from a clinical trial.     

Approximating the Non-Null Distribution of the Likelihood Ratio Test Statistic

This paper is concerned with investigation into the behavior of the likelihood ratio test statistic G² when the alternative hypothesis M (QQ) is the true model. Exact moments of G² are computed empirically and three approximations are considered for approximating the non-null distribution of G². Our results show that the two parameter gamma distribution provides a closer approximation to the exact powers of G². A randomized procedure was employed to obtain critical values based on 1000 simulated samples.     

Finite Sample Properties of the Maximum Likelihood Estimator and of Likelihood Ratio Tests in Hidden Markov Models

Hidden Markov models were successfully applied in various fields of time series analysis, especially for analyzing ion channel recordings. The maximum likelihood estimator (MLE) has recently been proven to be asymptotically normally distributed. Here, we investigate finite sample properties of the MLE and of different types of likelihood ratio tests (LRTs) by means of simulation studies. The MLE is shown to reach the asymptotic behavior within sample sizes that are common for various applications. Thus, reliable estimates and confidence intervals can be obtained. We give an approximative scaling function for the estimation error for finite samples, and investigate the power of different LRTs suitable for applications to ion channels, including tests for superimposed hidden Markov processes. Our results are applied to physiological sodium channel data.     

Use of the Maximum Likelihood Method in the Analysis of Phenotypic Stability

The stability variance is an important estimator of phenotypic stability of genotypes. It may be estimated by method of moments and by maximum likelihood. We demonstrate by Monte Carlo simulation that, given a sufficient number of environments, maximum likelihood estimates (MLE's) are slightly better if ranking of genotypes is the experimenter's major aim. A likelihood ratio test is available for different hypotheses.     

Likelihood ratio testing for admixture models with application to genetic linkage analysis

Summary We consider likelihood ratio tests (LRT) and their modifications for homogeneity in admixture models. The admixture model is a two‐component mixture model, where one component is indexed by an unknown parameter while the parameter value for the other component is known. This model is widely used in genetic linkage analysis under heterogeneity in which the kernel distribution is binomial. For such models, it is long recognized that testing for homogeneity is nonstandard, and the LRT statistic does not converge to a conventional χ² distribution. In this article, we investigate the asymptotic behavior of the LRT for general admixture models and show that its limiting distribution is equivalent to the supremum of a squared Gaussian process. We also discuss the connection and comparison between LRT and alternative approaches such as modifications of LRT and score tests, including the modified LRT ( Fu, Chen, and Kalbfleisch, 2006 , Statistica Sinica 16 , 805–823). The LRT is an omnibus test that is powerful to detect general alternative hypotheses. In contrast, alternative approaches may be slightly more powerful to detect certain type of alternatives, but much less powerful for others. Our results are illustrated by simulation studies and an application to a genetic linkage study of schizophrenia.     

Measuring the relative information in allele-sharing linkage studies

In the context of allele-sharing methods, this article investigates ways of measuring the information in the marker data relative to the amount of information that would have been available if the identity-by-descent (IBD) process were known. Such measures are needed to decide whether new markers can substantially modify the evidence for excess sharing. We propose new measures that take advantage of the properties of the exponential model introduced by Kong and Cox (1997, American Journal of Human Genetics61, 1179-1188). These measures are related to Fisher Information and hence are also efficiency measures. Large-sample and small-sample properties of the new and previously proposed measures of information are examined.     

Small Sample Distribution and Power of the Binomial Index of Dispersion and Log Likelihood Ratio Tests

This paper considers the exact distribution of the X² index of dispersion and -2 log (likelihood ratio) tests for the hypothesis of homogeneity of c independent samples from a common binomial population. The exact significance levels and power of these tests under ‘logit’ alternatives are compared numerically for the cases: c = 3, 4, 5 and various sample sizes. n₁ = 5,10 for i = 1,…, c.     

Testing the non-unity of rate ratio under inverse sampling

Inverse sampling is considered to be a more appropriate sampling scheme than the usual binomial sampling scheme when subjects arrive sequentially, when the underlying response of interest is acute, and when maximum likelihood estimators of some epidemiologic indices are undefined. In this article, we study various statistics for testing non-unity rate ratios in case-control studies under inverse sampling. These include the Wald, unconditional score, likelihood ratio and conditional score statistics. Three methods (the asymptotic, conditional exact, and Mid-P methods) are adopted for P-value calculation. We evaluate the performance of different combinations of test statistics and P-value calculation methods in terms of their empirical sizes and powers via Monte Carlo simulation. In general, asymptotic score and conditional score tests are preferable for their actual type I error rates are well controlled around the pre-chosen nominal level, and their powers are comparatively the largest. The exact version of Wald test is recommended if one wants to control the actual type I error rate at or below the pre-chosen nominal level. If larger power is expected and fluctuation of sizes around the pre-chosen nominal level are allowed, then the Mid-P version of Wald test is a desirable alternative. We illustrate the methodologies with a real example from a heart disease study.     

The Evaluation and Interpretation of Cervical Cytology: Application of the Likelihood Ratio Concept

The Papanicolaou smear (Pap test), used for the detection and prevention of neoplastic lesions of the cervix, is known to have both false negative and false positive results. Proper handling of the diagnostic uncertainty resulting from these errors demands quantification of flaws. Traditionally, sensitivity, specificity and predictive values are used for that aim. In this study another approach is advocated, namely the use of the likelihood ratio. For cervical cytology this ratio is the quotient of the probability of a Pap class within the diseased population to the probability of that same Pap class within the non-diseased group. This approach enables the characterization of each Pap class separately, and is therefore much better for clinical interpretation of the result. It is also a superior approach for quality assessment.