Volume 13, Issue 3

We assess the accuracy of various state-of-the-art statistics and machine learning methods for reconstructing gene and protein regulatory networks in the context of circadian regulation. Our study draws on the increasing availability of gene expression and protein concentration time series for key circadian clock components in Arabidopsis thaliana . In addition, gene expression and protein concentration time series are simulated from a recently published regulatory network of the circadian clock in A. thaliana , in which protein and gene interactions are described by a Markov jump process based on Michaelis-Menten kinetics. We closely follow recent experimental protocols, including the entrainment of seedlings to different light-dark cycles and the knock-out of various key regulatory genes. Our study provides relative network reconstruction accuracy scores for a critical comparative performance evaluation, and sheds light on a series of highly relevant questions: it quantifies the influence of systematically missing values related to unknown protein concentrations and mRNA transcription rates, it investigates the dependence of the performance on the network topology and the degree of recurrency, it provides deeper insight into when and why non-linear methods fail to outperform linear ones, it offers improved guidelines on parameter settings in different inference procedures, and it suggests new hypotheses about the structure of the central circadian gene regulatory network in A. thaliana .

This paper uses a Bayesian formulation of a clustering procedure to identify gene-gene interactions under case-control studies, called the Algorithm via Bayesian Clustering to Detect Epistasis (ABCDE). The ABCDE uses Dirichlet process mixtures to model SNP marker partitions, and uses the Gibbs weighted Chinese restaurant sampling to simulate posterior distributions of these partitions. Unlike the representative Bayesian epistasis detection algorithm BEAM, which partitions markers into three groups, the ABCDE can be evaluated at any given partition, regardless of the number of groups. This study also develops permutation tests to validate the disease association for SNP subsets identified by the ABCDE, which can yield results that are more robust to model specification and prior assumptions. This study examines the performance of the ABCDE and compares it with the BEAM using various simulated data and a schizophrenia SNP dataset.

We present a novel characterization of the generalized family wise error rate: k FWER. The interpretation allows researchers to view k FWER as a function of the test statistics rather than current methods based on p -values. Using this interpretation we present several theorems and methods (parametric and non-parametric) for estimating k FWER in various data settings. With this version of k FWER, researchers will have an estimate of k FWER in addition to knowing what tests are significant at the estimated k FWER. Additionally, we present methods that use empirical null distributions in place of parametric distributions in standard p -value k FWER controlling schemes. These advancements represent an improvement over common k FWER methods which are based on parametric assumptions and merely report the tests that are significant under a given value for k FWER.

Background: Identifying differentially expressed genes has been an important and widely used approach to investigate gene functions and molecular mechanisms. A related issue that has drawn much less attention but is equally important is the identification of constantly expressed genes across different conditions. A common practice is to treat genes that are not significantly differentially expressed as significantly equivalently expressed. Such naive practice often leads to large false discovery rate and low power. The more appropriate way for identifying constantly expressed genes should be conducting high dimensional statistical equivalence tests. A well-known equivalence test, the two one-sided tests (TOST), can be used for this purpose. However, due to the small sample sizes often associated with genomics data, the variance estimator in the TOST test could be unstable. Hence it would be fitting to examine the application of shrinkage variance estimators to the TOST test in high dimensional settings. Result: In this paper, we study the effect of shrinking the variance estimators in the TOST test in high dimensional settings through simulation studies. In addition, we derive analytic formulas for the p-value of the resultant shrinkage variance TOST test and apply it to a real data set.

Risk prediction models can link high-dimensional molecular measurements, such as DNA methylation, to clinical endpoints. For biological interpretation, often a sparse fit is desirable. Different molecular aggregation levels, such as considering DNA methylation at the CpG, gene, or chromosome level, might demand different degrees of sparsity. Hence, model building and estimation techniques should be able to adapt their sparsity according to the setting. Additionally, underestimation of coefficients, which is a typical problem of sparse techniques, should also be addressed. We propose a comprehensive approach, based on a boosting technique that allows a flexible adaptation of model sparsity and addresses these problems in an integrative way. The main motivation is to have an automatic sparsity adaptation. In a simulation study, we show that this approach reduces underestimation in sparse settings and selects more adequate model sizes than the corresponding non-adaptive boosting technique in non-sparse settings. Using different aggregation levels of DNA methylation data from a study in kidney carcinoma patients, we illustrate how automatically selected values of the sparsity tuning parameter can reflect the underlying structure of the data. In addition to that, prediction performance and variable selection stability is compared to the non-adaptive boosting approach.

Effective analytical tools are highly desirable for data analysis and for making the biological link between genotypic and phenotypic measures. In family data it is important to reconcile the methods that explain the phenotypic variability through fixed genetic effects and ones that estimate variance components using classical heritability methods. Thus, in this paper, we propose a method based on added-variable plot for polygenic linear mixed models applied to genome wide association studies in family-based designs. Our goal is to be able to discriminate genetic predictor variables in effects due to random polygenic and residual components. We also propose an index to detect influential families for each predictor variable identified with genetic effect. We assess the performance of our proposed method using our own family simulated data and the Genetic Analysis Workshop 17 family simulated data.

Stochastic systems in biology often exhibit substantial variability within and between cells. This variability, as well as having dramatic functional consequences, provides information about the underlying details of the system’s behavior. It is often desirable to infer properties of the parameters governing such systems given experimental observations of the mean and variance of observed quantities. In some circumstances, analytic forms for the likelihood of these observations allow very efficient inference: we present these forms and demonstrate their usage. When likelihood functions are unavailable or difficult to calculate, we show that an implementation of approximate Bayesian computation (ABC) is a powerful tool for parametric inference in these systems. However, the calculations required to apply ABC to these systems can also be computationally expensive, relying on repeated stochastic simulations. We propose an ABC approach that cheaply eliminates unimportant regions of parameter space, by addressing computationally simple mean behavior before explicitly simulating the more computationally demanding variance behavior. We show that this approach leads to a substantial increase in speed when applied to synthetic and experimental datasets.