Bayes and empirical-Bayes multiplicity adjustment in the variable-selection problem

This paper studies the multiplicity-correction effect of standard Bayesian variable-selection priors in linear regression. Our first goal is to clarify when, and how, multiplicity correction happens automatically in Bayesian analysis, and to distinguish this correction from the Bayesian Ockham's-razor effect. Our second goal is to contrast empirical-Bayes and fully Bayesian approaches to variable selection through examples, theoretical results and simulations. Considerable differences between the two approaches are found. In particular, we prove a theorem that characterizes a surprising aymptotic discrepancy between fully Bayes and empirical Bayes. This discrepancy arises from a different source than the failure to account for hyperparameter uncertainty in the empirical-Bayes estimate. Indeed, even at the extreme, when the empirical-Bayes estimate converges asymptotically to the true variable-inclusion probability, the potential for a serious difference remains. © Institute of Mathematical Statistics, 2010.

Missing data in randomized clinical trials for weight loss: scope of the problem, state of the field, and performance of statistical methods.

BACKGROUND: Dropouts and missing data are nearly-ubiquitous in obesity randomized controlled trails, threatening validity and generalizability of conclusions. Herein, we meta-analytically evaluate the extent of missing data, the frequency with which various analytic methods are employed to accommodate dropouts, and the performance of multiple statistical methods. METHODOLOGY/PRINCIPAL FINDINGS: We searched PubMed and Cochrane databases (2000-2006) for articles published in English and manually searched bibliographic references. Articles of pharmaceutical randomized controlled trials with weight loss or weight gain prevention as major endpoints were included. Two authors independently reviewed each publication for inclusion. 121 articles met the inclusion criteria. Two authors independently extracted treatment, sample size, drop-out rates, study duration, and statistical method used to handle missing data from all articles and resolved disagreements by consensus. In the meta-analysis, drop-out rates were substantial with the survival (non-dropout) rates being approximated by an exponential decay curve (e(-lambdat)) where lambda was estimated to be .0088 (95% bootstrap confidence interval: .0076 to .0100) and t represents time in weeks. The estimated drop-out rate at 1 year was 37%. Most studies used last observation carried forward as the primary analytic method to handle missing data. We also obtained 12 raw obesity randomized controlled trial datasets for empirical analyses. Analyses of raw randomized controlled trial data suggested that both mixed models and multiple imputation performed well, but that multiple imputation may be more robust when missing data are extensive. CONCLUSION/SIGNIFICANCE: Our analysis offers an equation for predictions of dropout rates useful for future study planning. Our raw data analyses suggests that multiple imputation is better than other methods for handling missing data in obesity randomized controlled trials, followed closely by mixed models. We suggest these methods supplant last observation carried forward as the primary method of analysis.

An approach to the study of G-protein-coupled receptor kinases: an in vitro-purified membrane assay reveals differential receptor specificity and regulation by G beta gamma subunits.

Phosphorylation of GTP-binding-regulatory (G)-protein-coupled receptors by specific G-protein-coupled receptor kinases (GRKs) is a major mechanism responsible for agonist-mediated desensitization of signal transduction processes. However, to date, studies of the specificity of these enzymes have been hampered by the difficulty of preparing the purified and reconstituted receptor preparations required as substrates. Here we describe an approach that obviates this problem by utilizing highly purified membrane preparations from Sf9 and 293 cells overexpressing G-protein-coupled receptors. We use this technique to demonstrate specificity of several GRKs with respect to both receptor substrates and the enhancing effects of G-protein beta gamma subunits on phosphorylation. Enriched membrane preparations of the beta 2- and alpha 2-C2-adrenergic receptors (ARs, where alpha 2-C2-AR refers to the AR whose gene is located on human chromosome 2) prepared by sucrose density gradient centrifugation from Sf9 or 293 cells contain the receptor at 100-300 pmol/mg of protein and serve as efficient substrates for agonist-dependent phosphorylation by beta-AR kinase 1 (GRK2), beta-AR kinase 2 (GRK3), or GRK5. Stoichiometries of agonist-mediated phosphorylation of the receptors by GRK2 (beta-AR kinase 1), in the absence and presence of G beta gamma, are 1 and 3 mol/mol, respectively. The rate of phosphorylation of the membrane receptors is 3 times faster than that of purified and reconstituted receptors. While phosphorylation of the beta 2-AR by GRK2, -3, and -5 is similar, the activity of GRK2 and -3 is enhanced by G beta gamma whereas that of GRK5 is not. In contrast, whereas GRK2 and -3 efficiently phosphorylate alpha 2-C2-AR, GRK5 is quite weak. The availability of a simple direct phosphorylation assay applicable to any cloned G-protein-coupled receptor should greatly facilitate elucidation of the mechanisms of regulation of these receptors by the expanding family of GRKs.

Communications inspired linear discriminant analysis

We study the problem of supervised linear dimensionality reduction, taking an information-theoretic viewpoint. The linear projection matrix is designed by maximizing the mutual information between the projected signal and the class label. By harnessing a recent theoretical result on the gradient of mutual information, the above optimization problem can be solved directly using gradient descent, without requiring simplification of the objective function. Theoretical analysis and empirical comparison are made between the proposed method and two closely related methods, and comparisons are also made with a method in which Rényi entropy is used to define the mutual information (in this case the gradient may be computed simply, under a special parameter setting). Relative to these alternative approaches, the proposed method achieves promising results on real datasets. Copyright 2012 by the author(s)/owner(s).

Atherosclerosis: Viewing the problem from a different perspective including possible treatment options

This paper proposes that atherosclerosis is initiated by a signaling event that deposits calcium hydroxyapatite (Ca-HAP). This event is preceded by a loss of mechanical structure in the arterial wall. After Ca-HAP has been deposited, it is unlikely that it will be reabsorbed because the solubility product constant (K sp) is very small, and the large stores of Ca +2 and PO 4-3 in the bones oppose any attempts to dissolve Ca-HAP by decreasing the common ions. The hydroxide ion (OH -) of Ca-HAP can be displaced in nature by fluoride (F -) and carbonate (CO 3-2) ions, and it is proposed that anions associated with cholesterol ester hydrolysis and, in very small quantities, the enolate of 7-ketocholesterol could also displace the OH -of Ca-HAP, forming an ionic bond. The free energy of hydration of Ca-HAP at 310 K is most likely negative, and the ionic radii of the anions associated with the hydrolysis of cholesterol ester are compatible with the substitution. Furthermore, examination of the pathology of atherosclerotic lesions by Raman and NMR spectroscopy and confocal microscopy supports deposition of Ca-HAP associated with cholesterol. Investigating the affinity of intermediates of cholesterol hydrolysis for Ca-HAP compared to lipoproteins such as HDL, LDL, and VLDL using isothermic titration calorimetry could add proof of this concept and may lead to the development of a new class of medications targeted at the deposition of cholesterol within Ca-HAP. Treatment of acute ischemic events as a consequence of atherosclerosis with denitrogenation and oxygenation is discussed. © the author(s), publisher and licensee Libertas Academica Ltd.

Component-wise gradient boosting and false discovery control in survival analysis with high-dimensional covariates.

MOTIVATION: Technological advances that allow routine identification of high-dimensional risk factors have led to high demand for statistical techniques that enable full utilization of these rich sources of information for genetics studies. Variable selection for censored outcome data as well as control of false discoveries (i.e. inclusion of irrelevant variables) in the presence of high-dimensional predictors present serious challenges. This article develops a computationally feasible method based on boosting and stability selection. Specifically, we modified the component-wise gradient boosting to improve the computational feasibility and introduced random permutation in stability selection for controlling false discoveries. RESULTS: We have proposed a high-dimensional variable selection method by incorporating stability selection to control false discovery. Comparisons between the proposed method and the commonly used univariate and Lasso approaches for variable selection reveal that the proposed method yields fewer false discoveries. The proposed method is applied to study the associations of 2339 common single-nucleotide polymorphisms (SNPs) with overall survival among cutaneous melanoma (CM) patients. The results have confirmed that BRCA2 pathway SNPs are likely to be associated with overall survival, as reported by previous literature. Moreover, we have identified several new Fanconi anemia (FA) pathway SNPs that are likely to modulate survival of CM patients. AVAILABILITY AND IMPLEMENTATION: The related source code and documents are freely available at https://sites.google.com/site/bestumich/issues. CONTACT: yili@umich.edu.