Zheyang Wu

Statistical Power of Model Selection Strategies for Genome-Wide Association Studies

Genome-wide association studies (GWAS) aim to identify genetic variants related to diseases by examining the associations between phenotypes and hundreds of thousands of genotyped markers. Because many genes are potentially involved in common diseases and a large number of markers are analyzed, it is crucial to devise an effective strategy to identify truly associated variants that have individual and/or interactive effects, while controlling false positives at the desired level. Although a number of model selection methods have been proposed in the literature, including marginal search, exhaustive search, and forward search, their relative performance has only been evaluated through limited simulations due to the lack of an analytical approach to calculating the power of these methods. This article develops a novel statistical approach for power calculation, derives accurate formulas for the power of different model selection strategies, and then uses the formulas to evaluate and compare these strategies in genetic model spaces. In contrast to previous studies, our theoretical framework allows for random genotypes, correlations among test statistics, and a false-positive control based on GWAS practice. After the accuracy of our analytical results is validated through simulations, they are utilized to systematically evaluate and compare the performance of these strategies in a wide class of genetic models. For a specific genetic model, our results clearly reveal how different factors, such as effect size, allele frequency, and interaction, jointly affect the statistical power of each strategy. An example is provided for the application of our approach to empirical research. The statistical approach used in our derivations is general and can be employed to address the model selection problems in other random predictor settings. We have developed an R package markerSearchPower to implement our formulas, which can be downloaded from the Comprehensive R Archive Network (CRAN) or http://bioinformatics.med.yale.edu/group/.

Identification of novel microRNAs in Xenopus laevis metaphase II arrested eggs

Using a combination of deep sequencing and bioinformatics approach, we for the first time identify miRNAs and their relative abundance in mature, metaphase II arrested eggs in Xenopus laevis. We characterize 115 miRNAs that have been described either in Xenopus tropicalis (85), X. laevis (9), or other vertebrate species (21) that also map to known Xenopus pre‐miRNAs and to the X. tropicalis genome. In addition, 72 new X. laevis putative candidate miRNAs are identified based on mapping to X. tropicalis genome within regions that have the propensity to form hairpin loops. These data expand on the availability of genetic information in X. laevis and identify target miRNAs for future functional studies. genesis 50:286–299, 2012.

A pathway analysis applied to Genetic Analysis Workshop 16 genome-wide rheumatoid arthritis data

The identification of several hundred genomic regions affecting disease risk has proven the ability of genome-wide association studies have proven their ability to identify genetic contributors to disease. Currently, single-nucleotide polymorphism (SNP) association analysis is the most widely used method of genome-wide association data, but recent research shows that multi-marker tests of association may provide greater power, especially when more than one mutation is present within a gene and the mutations are in low linkage disequilibrium with each other. Here we use a multi-marker association test based on regression to SNPs located within known genes to obtain a gene-level score of association. We then perform pathway analysis using this score as a measure of gene importance. We use two tests of pathway enrichment - a binomial test and a random set method. By utilizing publicly available gene and pathway information, we identify B cell, cytokine and inflammation response, and antigen presentation pathways as being associated with rheumatoid arthritis. These results confirm known biological mechanisms for auto-immunity disorders, of which rheumatoid arthritis is one.

Mechanical stress is associated with right ventricular response to pulmonary valve replacement in patients with repaired tetralogy of Fallot

OBJECTIVE Patients with repaired tetralogy of Fallot account for a substantial proportion of cases with late-onset right ventricular failure. The current surgical approach, which includes pulmonary valve replacement/insertion, has yielded mixed results. Therefore, it may be clinically useful to identify parameters that can be used to predict right ventricular function response to pulmonary valve replacement. METHODS Cardiac magnetic resonance data before and 6 months after pulmonary valve replacement were obtained from 16 patients with repaired tetralogy of Fallot (8 male, 8 female; median age, 42.75 years). Right ventricular ejection fraction change from pre- to postpulmonary valve replacement was used as the outcome. The patients were divided into group 1 (n = 8, better outcome) and group 2 (n = 8, worst outcome). Cardiac magnetic resonance-based patient-specific computational right ventricular/left ventricular models were constructed, and right ventricular mechanical stress and strain, wall thickness, curvature, and volumes were obtained for analysis. RESULTS Our results indicated that right ventricular wall stress was the best single predictor for postpulmonary valve replacement outcome with an area under the receiver operating characteristic curve of 0.819. Mean values of stress, strain, wall thickness, and longitudinal curvature differed significantly between the 2 groups with right ventricular wall stress showing the largest difference. Mean right ventricular stress in group 2 was 103% higher than in group 1. CONCLUSIONS Computational modeling and right ventricular stress may be used as tools to identify right ventricular function response to pulmonary valve replacement. Large-scale clinical studies are needed to validate these preliminary findings.

The Impact of Medicaid Coverage and Reimbursement on Access to Diagnostic Mammography

Women of low socioeconomic status are at risk for delayed evaluation of abnormal mammograms and later stage presentations of breast cancer. Medicaid reimbursement for clinical services is lower than Medicare reimbursement, yet it is unclear whether low Medicaid reimbursement is a barrier to accessing mammography. The objective of the current study was to determine the association between reported insurance type (Medicaid vs Medicare), Medicaid reimbursement rate, and access to diagnostic mammography (DM).

Longitudinal Data Analysis for Genetic Studies in the Whole-Genome Sequencing Era

The analysis of whole‐genome sequence (WGS) data using longitudinal phenotypes offers a potentially rich resource for the examination of the genetic variants and their covariates that affect complex phenotypes over time. We summarize eight contributions to the Genetic Analysis Workshop 18, which applied a diverse array of statistical genetic methods to analyze WGS data in combination with data from genome‐wide association studies (GWAS) from up to four different time points on blood pressure phenotypes. The common goal of these analyses was to develop and apply appropriate methods that utilize longitudinal repeated measures to potentially increase the analytic efficiency of WGS and GWAS data. These diverse methods can be grouped into two categories, based on the way they model dependence structures: (1) linear mixed‐effects (LME) models, where the random effect terms in the linear models are used to capture the dependence structures; and (2) variance‐components models, where the dependence structures are constructed directly based on multiple components of variance‐covariance matrices for the multivariate Gaussian responses. Despite the heterogeneous nature of these analytical methods, the group came to the following conclusions: (1) the use of repeat measurements can gain power to identify variants associated with the phenotype; (2) the inclusion of family data may correct genotyping errors and allow for more accurate detection of rare variants than using unrelated individuals only; and (3) fitting mixed‐effects and variance‐components models for longitudinal data presents computational challenges. The challenges and computational burden demanded by WGS data were addressed in the eight contributions.

Morphological and Stress Vulnerability Indices for Human Coronary Plaques and Their Correlations with Cap Thickness and Lipid Percent: An IVUS-Based Fluid-Structure Interaction Multi-Patient Study

Plaque vulnerability, defined as the likelihood that a plaque would rupture, is difficult to quantify due to lack of in vivo plaque rupture data. Morphological and stress-based plaque vulnerability indices were introduced as alternatives to obtain quantitative vulnerability assessment. Correlations between these indices and key plaque features were investigated. In vivo intravascular ultrasound (IVUS) data were acquired from 14 patients and IVUS-based 3D fluid-structure interaction (FSI) coronary plaque models with cyclic bending were constructed to obtain plaque wall stress/strain and flow shear stress for analysis. For the 617 slices from the 14 patients, lipid percentage, min cap thickness, critical plaque wall stress (CPWS), strain (CPWSn) and flow shear stress (CFSS) were recorded, and cap index, lipid index and morphological index were assigned to each slice using methods consistent with American Heart Association (AHA) plaque classification schemes. A stress index was introduced based on CPWS. Linear Mixed-Effects (LME) models were used to analyze the correlations between the mechanical and morphological indices and key morphological factors associated with plaque rupture. Our results indicated that for all 617 slices, CPWS correlated with min cap thickness, cap index, morphological index with r = -0.6414, 0.7852, and 0.7411 respectively (p<0.0001). The correlation between CPWS and lipid percentage, lipid index were weaker (r = 0.2445, r = 0.2338, p<0.0001). Stress index correlated with cap index, lipid index, morphological index positively with r = 0.8185, 0.3067, and 0.7715, respectively, all with p<0.0001. For all 617 slices, the stress index has 66.77% agreement with morphological index. Morphological and stress indices may serve as quantitative plaque vulnerability assessment supported by their strong correlations with morphological features associated with plaque rupture. Differences between the two indices may lead to better plaque assessment schemes when both indices were jointly used with further validations from clinical studies.

Integrated multidimensional analysis is required for accurate prognostic biomarkers in colorectal cancer.

CRC cancer is one of the deadliest diseases in Western countries. In order to develop prognostic biomarkers for CRC (colorectal cancer) aggressiveness, we analyzed retrospectively 267 CRC patients via a novel, multidimensional biomarker platform. Using nanofluidic technology for qPCR analysis and quantitative fluorescent immunohistochemistry for protein analysis, we assessed 33 microRNAs, 124 mRNAs and 9 protein antigens. Analysis was conducted in each single dimension (microRNA, gene or protein) using both the multivariate Cox model and Kaplan-Meier method. Thereafter, we simplified the censored survival data into binary response data (aggressive vs. non aggressive cancer). Subsequently, we integrated the data into a diagnostic score using sliced inverse regression for sufficient dimension reduction. Accuracy was assessed using area under the receiver operating characteristic curve (AUC). Single dimension analysis led to the discovery of individual factors that were significant predictors of outcome. These included seven specific microRNAs, four genes, and one protein. When these factors were quantified individually as predictors of aggressive disease, the highest demonstrable area under the curve (AUC) was 0.68. By contrast, when all results from single dimensions were combined into integrated biomarkers, AUCs were dramatically increased with values approaching and even exceeding 0.9. Single dimension analysis generates statistically significant predictors, but their predictive strengths are suboptimal for clinical utility. A novel, multidimensional integrated approach overcomes these deficiencies. Newly derived integrated biomarkers have the potential to meaningfully guide the selection of therapeutic strategies for individual patients while elucidating molecular mechanisms driving disease progression.

Gene hunting of the Genetic Analysis Workshop 16 rheumatoid arthritis data using rough set theory

We propose to use the rough set theory to identify genes affecting rheumatoid arthritis risk from the data collected by the North American Rheumatoid Arthritis Consortium. For each gene, we employ generalized dynamic reducts in the rough set theory to select a subset of single-nucleotide polymorphisms (SNPs) to represent the genetic information from this gene. We then group the study subjects into different clusters based on their genotype similarity at the selected markers. Statistical association between disease status and cluster membership is then studied to identify genes associated with rheumatoid arthritis. Based on our proposed approach, we are able to identify a number of statistically significant genes associated with rheumatoid arthritis. Aside from genes on chromosome 6, our identified genes include known disease-associated genes such as PTPN22 and TRAF1. In addition, our list contains other biologically plausible genes, such as ADAM15 and AGPAT2. Our findings suggest that ADAM15 and AGPAT2 may contribute to a genetic predisposition through abnormal angiogenesis and adipose tissue.

Patient-Specific MRI-Based Right Ventricle Models Using Different Zero-Load Diastole and Systole Geometries for Better Cardiac Stress and Strain Calculations and Pulmonary Valve Replacement Surgical Outcome Predictions

Background Accurate calculation of ventricular stress and strain is critical for cardiovascular investigations. Sarcomere shortening in active contraction leads to change of ventricular zero-stress configurations during the cardiac cycle. A new model using different zero-load diastole and systole geometries was introduced to provide more accurate cardiac stress/strain calculations with potential to predict post pulmonary valve replacement (PVR) surgical outcome. Methods Cardiac magnetic resonance (CMR) data were obtained from 16 patients with repaired tetralogy of Fallot prior to and 6 months after pulmonary valve replacement (8 male, 8 female, mean age 34.5 years). Patients were divided into Group 1 (n = 8) with better post PVR outcome and Group 2 (n = 8) with worse post PVR outcome based on their change in RV ejection fraction (EF). CMR-based patient-specific computational RV/LV models using one zero-load geometry (1G model) and two zero-load geometries (diastole and systole, 2G model) were constructed and RV wall thickness, volume, circumferential and longitudinal curvatures, mechanical stress and strain were obtained for analysis. Pairwise T-test and Linear Mixed Effect (LME) model were used to determine if the differences from the 1G and 2G models were statistically significant, with the dependence of the pair-wise observations and the patient-slice clustering effects being taken into consideration. For group comparisons, continuous variables (RV volumes, WT, C- and L- curvatures, and stress and strain values) were summarized as mean ± SD and compared between the outcome groups by using an unpaired Student t-test. Logistic regression analysis was used to identify potential morphological and mechanical predictors for post PVR surgical outcome. Results Based on results from the 16 patients, mean begin-ejection stress and strain from the 2G model were 28% and 40% higher than that from the 1G model, respectively. Using the 2G model results, RV EF changes correlated negatively with stress (r = -0.609, P = 0.012) and with pre-PVR RV end-diastole volume (r = -0.60, P = 0.015), but did not correlate with WT, C-curvature, L-curvature, or strain. At begin-ejection, mean RV stress of Group 2 was 57.4% higher than that of Group 1 (130.1±60.7 vs. 82.7±38.8 kPa, P = 0.0042). Stress was the only parameter that showed significant differences between the two groups. The combination of circumferential curvature, RV volume and the difference between begin-ejection stress and end-ejection stress was the best predictor for post PVR outcome with an area under the ROC curve of 0.855. The begin-ejection stress was the best single predictor among the 8 individual parameters with an area under the ROC curve of 0.782. Conclusion The new 2G model may be able to provide more accurate ventricular stress and strain calculations for potential clinical applications. Combining morphological and mechanical parameters may provide better predictions for post PVR outcome.