Cytometry is an advanced technique for simultaneously identifying and quantifying many cell surface and intracellular proteins at a single-cell resolution. Analyzing high-dimensional cytometry data involves identifying and quantifying cell populations based on their marker expressions. This study provided a quantitative review and comparison of various ways to phenotype cellular populations within the cytometry data, including manual gating, unsupervised clustering, and supervised auto-gating. Six datasets from diverse species and sample types were included in the study, and manual gating with two hierarchical layers was used as the truth for evaluation. For manual gating, results from five researchers were compared to illustrate the gating consistency among different raters. For unsupervised clustering, 23 tools were quantitatively compared in terms of accuracy with the truth and computing cost. While no method outperformed all others, several tools, including PAC-MAN, CCAST, FlowSOM, flowClust, and DEPECHE, generally demonstrated strong performance. For supervised auto-gating methods, four algorithms were evaluated, where DeepCyTOF and CyTOF Linear Classifier performed the best. We further provided practical recommendations on prioritizing gating methods based on different application scenarios. This study offers comprehensive insights for biologists to understand diverse gating methods and choose the best-suited ones for their applications.

SUMMARY: Causal mediation analysis investigates the role of mediators in the relationship between exposure and outcome. In the analysis of omics or imaging data, mediators are often high-dimensional, presenting challenges such as multicollinearity and interpretability. Existing methods either compromise interpretability or fail to effectively prioritize mediators. To address these challenges and advance causal mediation analysis in high-dimensional contexts, we propose the Partial Sum Statistic and Sample Splitting Strategy (PS5) framework. Through extensive simulations, we demonstrate that PS5 offers superior type I error control, higher statistical power, reduced bias in mediation effect estimation, and more accurate mediator selection. We apply PS5 to an imaging genetics dataset of chronic obstructive pulmonary disease (COPD) patients from the COPDGene study. The results show successful estimation of the global indirect effect and identification of mediating image regions. Notably, we identify a region in the lower lobe of the lung that exhibits a strong and concordant mediation effect for both genetic and environmental exposures, suggesting potential targets for treatment to mitigate COPD severity caused by genetic and smoking effects. AVAILABILITY AND IMPLEMENTATION: PS5 is publicly available at https://github.com/hung-ching-chang/PS5Med.

MOTIVATION: Biomarker detection plays a pivotal role in biomedical research. Integrating omics studies from multiple cohorts can enhance statistical power, accuracy, and robustness of the detection results. However, existing methods for horizontally combining omics studies are mostly designed for two-class scenarios (e.g. cases versus controls) and are not directly applicable for studies with multi-class design (e.g. samples from multiple disease subtypes, treatments, tissues, or cell types). RESULTS: We propose a statistical framework, namely Mutual Information Concordance Analysis (MICA), to detect biomarkers with concordant multi-class expression pattern across multiple omics studies from an information theoretic perspective. Our approach first detects biomarkers with concordant multi-class patterns across partial or all of the omics studies using a global test by mutual information. A post hoc analysis is then performed for each detected biomarkers and identify studies with concordant pattern. Extensive simulations demonstrate improved accuracy and successful false discovery rate control of MICA compared to an existing multi-class correlation method. The method is then applied to two practical scenarios: four tissues of mouse metabolism-related transcriptomic studies, and three sources of estrogen treatment expression profiles. Detected biomarkers by MICA show intriguing biological insights and functional annotations. Additionally, we implemented MICA for single-cell RNA-Seq data for tumor progression biomarkers, highlighting critical roles of ribosomal function in the tumor microenvironment of triple-negative breast cancer and underscoring the potential of MICA for detecting novel therapeutic targets. AVAILABILITY AND IMPLEMENTATION: The source code is available on Figshare at https://doi.org/10.6084/m9.figshare.27635436. Additionally, the R package can be installed directly from GitHub at https://github.com/jianzou75/MICA.

Cancer models are instrumental as a substitute for human studies and to expedite basic, translational, and clinical cancer research. For a given cancer type, a wide selection of models, such as cell lines, patient-derived xenografts, organoids and genetically modified murine models, are often available to researchers. However, how to quantify their congruence to human tumors and to select the most appropriate cancer model is a largely unsolved issue. Here, we present Congruence Analysis and Selection of CAncer Models (CASCAM), a statistical and machine learning framework for authenticating and selecting the most representative cancer models in a pathway-specific manner using transcriptomic data. CASCAM provides harmonization between human tumor and cancer model omics data, systematic congruence quantification, and pathway-based topological visualization to determine the most appropriate cancer model selection. The systems approach is presented using invasive lobular breast carcinoma (ILC) subtype and suggesting CAMA1 followed by UACC3133 as the most representative cell lines for ILC research. Two additional case studies for triple negative breast cancer (TNBC) and patient-derived xenograft/organoid (PDX/PDO) are further investigated. CASCAM is generalizable to any cancer subtype and will authenticate cancer models for faithful non-human preclinical research towards precision medicine.

SUMMARY: Circadian oscillations of gene expression regulate daily physiological processes, and their disruption is linked to many diseases. Circadian rhythms can be disrupted in a variety of ways, including differential phase, amplitude and rhythm fitness. Although many differential circadian biomarker detection methods have been proposed, a workflow for systematic detection of multifaceted differential circadian characteristics with accurate false positive control is not currently available. We propose a comprehensive and interactive pipeline to capture the multifaceted characteristics of differentially rhythmic biomarkers. Analysis outputs are accompanied by informative visualization and interactive exploration. The workflow is demonstrated in multiple case studies and is extensible to general omics applications. AVAILABILITY AND IMPLEMENTATION: R package, Shiny app and source code are available in GitHub (https://github.com/DiffCircaPipeline) and Zenodo (https://doi.org/10.5281/zenodo.7507989). SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: The advancement of high-throughput technology characterizes a wide variety of epigenetic modifications and noncoding RNAs across the genome involved in disease pathogenesis via regulating gene expression. The high dimensionality of both epigenetic/noncoding RNA and gene expression data make it challenging to identify the important regulators of genes. Conducting univariate test for each possible regulator-gene pair is subject to serious multiple comparison burden, and direct application of regularization methods to select regulator-gene pairs is computationally infeasible. Applying fast screening to reduce dimension first before regularization is more efficient and stable than applying regularization methods alone. RESULTS: We propose a novel screening method based on robust partial correlation to detect epigenetic and noncoding RNA regulators of gene expression over the whole genome, a problem that includes both high-dimensional predictors and high-dimensional responses. Compared to existing screening methods, our method is conceptually innovative that it reduces the dimension of both predictor and response, and screens at both node (regulators or genes) and edge (regulator-gene pairs) levels. We develop data-driven procedures to determine the conditional sets and the optimal screening threshold, and implement a fast iterative algorithm. Simulations and applications to long noncoding RNA and microRNA regulation in Kidney cancer and DNA methylation regulation in Glioblastoma Multiforme illustrate the validity and advantage of our method. AVAILABILITY AND IMPLEMENTATION: The R package, related source codes and real datasets used in this article are provided at https://github.com/kehongjie/rPCor. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Circadian rhythmicity in transcriptomic profiles has been shown in many physiological processes, and the disruption of circadian patterns has been found to associate with several diseases. In this paper, we developed a series of likelihood-based methods to detect (i) circadian rhythmicity (denoted as LR_rhythmicity) and (ii) differential circadian patterns comparing two experimental conditions (denoted as LR_diff). In terms of circadian rhythmicity detection, we demonstrated that our proposed LR_rhythmicity could better control the type I error rate compared to existing methods under a wide variety of simulation settings. In terms of differential circadian patterns, we developed methods in detecting differential amplitude, differential phase, differential basal level and differential fit, which also successfully controlled the type I error rate. In addition, we demonstrated that the proposed LR_diff could achieve higher statistical power in detecting differential fit, compared to existing methods. The superior performance of LR_rhythmicity and LR_diff was demonstrated in four real data applications, including a brain aging data (gene expression microarray data of human postmortem brain), a time-restricted feeding data (RNA sequencing data of human skeletal muscles) and a scRNAseq data (single cell RNA sequencing data of mouse suprachiasmatic nucleus). An R package for our methods is publicly available on GitHub https://github.com/diffCircadian/diffCircadian.

Alternative polyadenylation (APA) in breast tumor samples results in the removal/addition of cis-regulatory elements such as microRNA (miRNA) target sites in the 3'-untranslated region (3'-UTRs) of genes. Although previous computational APA studies focused on a subset of genes strongly affected by APA (APA genes), we identify miRNAs of which widespread APA events collectively increase or decrease the number of target sites [probabilistic inference of microRNA target site modification through APA (PRIMATA-APA)]. Using PRIMATA-APA on the cancer genome atlas (TCGA) breast cancer data, we found that the global APA events change the number of the target sites of particular microRNAs [target sites modified miRNA (tamoMiRNA)] enriched for cancer development and treatments. We also found that when knockdown (KD) of NUDT21 in HeLa cells induces a different set of widespread 3'-UTR shortening than TCGA breast cancer data, it changes the target sites of the common tamoMiRNAs. Since the NUDT21 KD experiment previously demonstrated the tumorigenic role of APA events in a miRNA dependent fashion, this result suggests that the APA-initiated tumorigenesis is attributable to the miRNA target site changes, not the APA events themselves. Further, we found that the miRNA target site changes identify tumor cell proliferation and immune cell infiltration to the tumor microenvironment better than the miRNA expression levels or the APA events themselves. Altogether, our computational analyses provide a proof-of-concept demonstration that the miRNA target site information indicates the effect of global APA events with a potential as predictive biomarker.

MOTIVATION: There is growing interest in the biomedical research community to incorporate retrospective data, available in healthcare systems, to shed light on associations between different biomarkers. Understanding the association between various types of biomedical data, such as genetic, blood biomarkers, imaging, etc. can provide a holistic understanding of human diseases. To formally test a hypothesized association between two types of data in Electronic Health Records (EHRs), one requires a substantial sample size with both data modalities to achieve a reasonable power. Current association test methods only allow using data from individuals who have both data modalities. Hence, researchers cannot take advantage of much larger EHR samples that includes individuals with at least one of the data types, which limits the power of the association test. RESULTS: We present a new method called the Semi-paired Association Test (SAT) that makes use of both paired and unpaired data. In contrast to classical approaches, incorporating unpaired data allows SAT to produce better control of false discovery and to improve the power of the association test. We study the properties of the new test theoretically and empirically, through a series of simulations and by applying our method on real studies in the context of Chronic Obstructive Pulmonary Disease. We are able to identify an association between the high-dimensional characterization of Computed Tomography chest images and several blood biomarkers as well as the expression of dozens of genes involved in the immune system. AVAILABILITY AND IMPLEMENTATION: Code is available on https://github.com/batmanlab/Semi-paired-Association-Test. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: Meta-analysis methods have been widely used to combine results from multiple clinical or genomic studies to increase statistical powers and ensure robust and accurate conclusions. The adaptively weighted Fisher's method (AW-Fisher), initially developed for omics applications but applicable for general meta-analysis, is an effective approach to combine P-values from K independent studies and to provide better biological interpretability by characterizing which studies contribute to the meta-analysis. Currently, AW-Fisher suffers from the lack of fast P-value computation and variability estimate of AW weights. When the number of studies K is large, the 3K - 1 possible differential expression pattern categories generated by AW-Fisher can become intractable. In this paper, we develop an importance sampling scheme with spline interpolation to increase the accuracy and speed of the P-value calculation. We also apply bootstrapping to construct a variability index for the AW-Fisher weight estimator and a co-membership matrix to categorize (cluster) differentially expressed genes based on their meta-patterns for intuitive biological investigations. RESULTS: The superior performance of the proposed methods is shown in simulations as well as two real omics meta-analysis applications to demonstrate its insightful biological findings. AVAILABILITY AND IMPLEMENTATION: An R package AWFisher (calling C++) is available at Bioconductor and GitHub (https://github.com/Caleb-Huo/AWFisher), and all datasets and programing codes for this paper are available in the Supplementary Material. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

SUMMARY: The rapid advances of omics technologies have generated abundant genomic data in public repositories and effective analytical approaches are critical to fully decipher biological knowledge inside these data. Meta-analysis combines multiple studies of a related hypothesis to improve statistical power, accuracy and reproducibility beyond individual study analysis. To date, many transcriptomic meta-analysis methods have been developed, yet few thoughtful guidelines exist. Here, we introduce a comprehensive analytical pipeline and browser-based software suite, called MetaOmics, to meta-analyze multiple transcriptomic studies for various biological purposes, including quality control, differential expression analysis, pathway enrichment analysis, differential co-expression network analysis, prediction, clustering and dimension reduction. The pipeline includes many public as well as >10 in-house transcriptomic meta-analytic methods with data-driven and biological-aim-driven strategies, hands-on protocols, an intuitive user interface and step-by-step instructions. AVAILABILITY AND IMPLEMENTATION: MetaOmics is freely available at https://github.com/metaOmics/metaOmics. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Motivation: Integrative analysis of multi-omics data from different high-throughput experimental platforms provides valuable insight into regulatory mechanisms associated with complex diseases, and gains statistical power to detect markers that are otherwise overlooked by single-platform omics analysis. In practice, a significant portion of samples may not be measured completely due to insufficient tissues or restricted budget (e.g. gene expression profile are measured but not methylation). Current multi-omics integrative methods require complete data. A common practice is to ignore samples with any missing platform and perform complete case analysis, which leads to substantial loss of statistical power. Methods: In this article, inspired by the popular Integrative Bayesian Analysis of Genomics data (iBAG), we propose a full Bayesian model that allows incorporation of samples with missing omics data. Results: Simulation results show improvement of the new full Bayesian approach in terms of outcome prediction accuracy and feature selection performance when sample size is limited and proportion of missingness is large. When sample size is large or the proportion of missingness is low, incorporating samples with missingness may introduce extra inference uncertainty and generate worse prediction and feature selection performance. To determine whether and how to incorporate samples with missingness, we propose a self-learning cross-validation (CV) decision scheme. Simulations and a real application on child asthma dataset demonstrate superior performance of the CV decision scheme when various types of missing mechanisms are evaluated. Availability and implementation: Freely available on the GitHub at https://github.com/CHPGenetics/FBM. Supplementary information: Supplementary data are available at Bioinformatics online.

Motivation: With the prevalent usage of microarray and massively parallel sequencing, numerous high-throughput omics datasets have become available in the public domain. Integrating abundant information among omics datasets is critical to elucidate biological mechanisms. Due to the high-dimensional nature of the data, methods such as principal component analysis (PCA) have been widely applied, aiming at effective dimension reduction and exploratory visualization. Results: In this article, we combine multiple omics datasets of identical or similar biological hypothesis and introduce two variations of meta-analytic framework of PCA, namely MetaPCA. Regularization is further incorporated to facilitate sparse feature selection in MetaPCA. We apply MetaPCA and sparse MetaPCA to simulations, three transcriptomic meta-analysis studies in yeast cell cycle, prostate cancer, mouse metabolism and a TCGA pan-cancer methylation study. The result shows improved accuracy, robustness and exploratory visualization of the proposed framework. Availability and implementation: An R package MetaPCA is available online. (http://tsenglab.biostat.pitt.edu/software.htm). Contact: [email protected]. Supplementary information: Supplementary data are available at Bioinformatics online.

MOTIVATION: Although coexpression analysis via pair-wise expression correlation is popularly used to elucidate gene-gene interactions at the whole-genome scale, many complicated multi-gene regulations require more advanced detection methods. Liquid association (LA) is a powerful tool to detect the dynamic correlation of two gene variables depending on the expression level of a third variable (LA scouting gene). LA detection from single transcriptomic study, however, is often unstable and not generalizable due to cohort bias, biological variation and limited sample size. With the rapid development of microarray and NGS technology, LA analysis combining multiple gene expression studies can provide more accurate and stable results. RESULTS: In this article, we proposed two meta-analytic approaches for LA analysis (MetaLA and MetaMLA) to combine multiple transcriptomic studies. To compensate demanding computing, we also proposed a two-step fast screening algorithm for more efficient genome-wide screening: bootstrap filtering and sign filtering. We applied the methods to five Saccharomyces cerevisiae datasets related to environmental changes. The fast screening algorithm reduced 98% of running time. When compared with single study analysis, MetaLA and MetaMLA provided stronger detection signal and more consistent and stable results. The top triplets are highly enriched in fundamental biological processes related to environmental changes. Our method can help biologists understand underlying regulatory mechanisms under different environmental exposure or disease states. AVAILABILITY AND IMPLEMENTATION: A MetaLA R package, data and code for this article are available at http://tsenglab.biostat.pitt.edu/software.htm. CONTACT: [email protected]. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: Supervised machine learning is widely applied to transcriptomic data to predict disease diagnosis, prognosis or survival. Robust and interpretable classifiers with high accuracy are usually favored for their clinical and translational potential. The top scoring pair (TSP) algorithm is an example that applies a simple rank-based algorithm to identify rank-altered gene pairs for classifier construction. Although many classification methods perform well in cross-validation of single expression profile, the performance usually greatly reduces in cross-study validation (i.e. the prediction model is established in the training study and applied to an independent test study) for all machine learning methods, including TSP. The failure of cross-study validation has largely diminished the potential translational and clinical values of the models. The purpose of this article is to develop a meta-analytic top scoring pair (MetaKTSP) framework that combines multiple transcriptomic studies and generates a robust prediction model applicable to independent test studies. RESULTS: We proposed two frameworks, by averaging TSP scores or by combining P-values from individual studies, to select the top gene pairs for model construction. We applied the proposed methods in simulated data sets and three large-scale real applications in breast cancer, idiopathic pulmonary fibrosis and pan-cancer methylation. The result showed superior performance of cross-study validation accuracy and biomarker selection for the new meta-analytic framework. In conclusion, combining multiple omics data sets in the public domain increases robustness and accuracy of the classification model that will ultimately improve disease understanding and clinical treatment decisions to benefit patients. AVAILABILITY AND IMPLEMENTATION: An R package MetaKTSP is available online. (http://tsenglab.biostat.pitt.edu/software.htm). CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: As sequencing technologies can help researchers detect common and rare variants across the human genome in many individuals, it is known that jointly calling genotypes across multiple individuals based on linkage disequilibrium (LD) can facilitate the analysis of low to modest coverage sequence data. However, genotype-calling methods for family-based sequence data, particularly for complex families beyond parent-offspring trios, are still lacking. RESULTS: In this study, first, we proposed an algorithm that considers both linkage disequilibrium (LD) patterns and familial transmission in nuclear and multi-generational families while retaining the computational efficiency. Second, we extended our method to incorporate external reference panels to analyze family-based sequence data with a small sample size. In simulation studies, we show that modeling multiple offspring can dramatically increase genotype calling accuracy and reduce phasing and Mendelian errors, especially at low to modest coverage. In addition, we show that using external panels can greatly facilitate genotype calling of sequencing data with a small number of individuals. We applied our method to a whole genome sequencing study of 1339 individuals at ~10X coverage from the Minnesota Center for Twin and Family Research. CONCLUSIONS: The aggregated results show that our methods significantly outperform existing ones that ignore family constraints or LD information. We anticipate that our method will be useful for many ongoing family-based sequencing projects. We have implemented our methods efficiently in a C++ program FamLDCaller, which is available from http://www.pitt.edu/~wec47/famldcaller.html.

MOTIVATION: Supervised machine learning is commonly applied in genomic research to construct a classifier from the training data that is generalizable to predict independent testing data. When test datasets are not available, cross-validation is commonly used to estimate the error rate. Many machine learning methods are available, and it is well known that no universally best method exists in general. It has been a common practice to apply many machine learning methods and report the method that produces the smallest cross-validation error rate. Theoretically, such a procedure produces a selection bias. Consequently, many clinical studies with moderate sample sizes (e.g. n = 30-60) risk reporting a falsely small cross-validation error rate that could not be validated later in independent cohorts. RESULTS: In this article, we illustrated the probabilistic framework of the problem and explored the statistical and asymptotic properties. We proposed a new bias correction method based on learning curve fitting by inverse power law (IPL) and compared it with three existing methods: nested cross-validation, weighted mean correction and Tibshirani-Tibshirani procedure. All methods were compared in simulation datasets, five moderate size real datasets and two large breast cancer datasets. The result showed that IPL outperforms the other methods in bias correction with smaller variance, and it has an additional advantage to extrapolate error estimates for larger sample sizes, a practical feature to recommend whether more samples should be recruited to improve the classifier and accuracy. An R package 'MLbias' and all source files are publicly available. AVAILABILITY AND IMPLEMENTATION: tsenglab.biostat.pitt.edu/software.htm. CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: In modern biomedical research of complex diseases, a large number of demographic and clinical variables, herein called phenomic data, are often collected and missing values (MVs) are inevitable in the data collection process. Since many downstream statistical and bioinformatics methods require complete data matrix, imputation is a common and practical solution. In high-throughput experiments such as microarray experiments, continuous intensities are measured and many mature missing value imputation methods have been developed and widely applied. Numerous methods for missing data imputation of microarray data have been developed. Large phenomic data, however, contain continuous, nominal, binary and ordinal data types, which void application of most methods. Though several methods have been developed in the past few years, not a single complete guideline is proposed with respect to phenomic missing data imputation. RESULTS: In this paper, we investigated existing imputation methods for phenomic data, proposed a self-training selection (STS) scheme to select the best imputation method and provide a practical guideline for general applications. We introduced a novel concept of "imputability measure" (IM) to identify missing values that are fundamentally inadequate to impute. In addition, we also developed four variations of K-nearest-neighbor (KNN) methods and compared with two existing methods, multivariate imputation by chained equations (MICE) and missForest. The four variations are imputation by variables (KNN-V), by subjects (KNN-S), their weighted hybrid (KNN-H) and an adaptively weighted hybrid (KNN-A). We performed simulations and applied different imputation methods and the STS scheme to three lung disease phenomic datasets to evaluate the methods. An R package "phenomeImpute" is made publicly available. CONCLUSIONS: Simulations and applications to real datasets showed that MICE often did not perform well; KNN-A, KNN-H and random forest were among the top performers although no method universally performed the best. Imputation of missing values with low imputability measures increased imputation errors greatly and could potentially deteriorate downstream analyses. The STS scheme was accurate in selecting the optimal method by evaluating methods in a second layer of missingness simulation. All source files for the simulation and the real data analyses are available on the author's publication website.

BACKGROUND: As high-throughput genomic technologies become accurate and affordable, an increasing number of data sets have been accumulated in the public domain and genomic information integration and meta-analysis have become routine in biomedical research. In this paper, we focus on microarray meta-analysis, where multiple microarray studies with relevant biological hypotheses are combined in order to improve candidate marker detection. Many methods have been developed and applied in the literature, but their performance and properties have only been minimally investigated. There is currently no clear conclusion or guideline as to the proper choice of a meta-analysis method given an application; the decision essentially requires both statistical and biological considerations. RESULTS: We performed 12 microarray meta-analysis methods for combining multiple simulated expression profiles, and such methods can be categorized for different hypothesis setting purposes: (1) HS(A): DE genes with non-zero effect sizes in all studies, (2) HS(B): DE genes with non-zero effect sizes in one or more studies and (3) HS(r): DE gene with non-zero effect in "majority" of studies. We then performed a comprehensive comparative analysis through six large-scale real applications using four quantitative statistical evaluation criteria: detection capability, biological association, stability and robustness. We elucidated hypothesis settings behind the methods and further apply multi-dimensional scaling (MDS) and an entropy measure to characterize the meta-analysis methods and data structure, respectively. CONCLUSIONS: The aggregated results from the simulation study categorized the 12 methods into three hypothesis settings (HS(A), HS(B), and HS(r)). Evaluation in real data and results from MDS and entropy analyses provided an insightful and practical guideline to the choice of the most suitable method in a given application. All source files for simulation and real data are available on the author's publication website.

SUMMARY: With the rapid advances and prevalence of high-throughput genomic technologies, integrating information of multiple relevant genomic studies has brought new challenges. Microarray meta-analysis has become a frequently used tool in biomedical research. Little effort, however, has been made to develop a systematic pipeline and user-friendly software. In this article, we present MetaOmics, a suite of three R packages MetaQC, MetaDE and MetaPath, for quality control, differentially expressed gene identification and enriched pathway detection for microarray meta-analysis. MetaQC provides a quantitative and objective tool to assist study inclusion/exclusion criteria for meta-analysis. MetaDE and MetaPath were developed for candidate marker and pathway detection, which provide choices of marker detection, meta-analysis and pathway analysis methods. The system allows flexible input of experimental data, clinical outcome (case-control, multi-class, continuous or survival) and pathway databases. It allows missing values in experimental data and utilizes multi-core parallel computing for fast implementation. It generates informative summary output and visualization plots, operates on different operation systems and can be expanded to include new algorithms or combine different types of genomic data. This software suite provides a comprehensive tool to conveniently implement and compare various genomic meta-analysis pipelines. AVAILABILITY: http://www.biostat.pitt.edu/bioinfo/software.htm CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: Detecting candidate markers in transcriptomic studies often encounters difficulties in complex diseases, particularly when overall signals are weak and sample size is small. Covariates including demographic, clinical and technical variables are often confounded with the underlying disease effects, which further hampers accurate biomarker detection. Our motivating example came from an analysis of five microarray studies in major depressive disorder (MDD), a heterogeneous psychiatric illness with mostly uncharacterized genetic mechanisms. RESULTS: We applied a random intercept model to account for confounding variables and case-control paired design. A variable selection scheme was developed to determine the effective confounders in each gene. Meta-analysis methods were used to integrate information from five studies and post hoc analyses enhanced biological interpretations. Simulations and application results showed that the adjustment for confounding variables and meta-analysis improved detection of biomarkers and associated pathways. CONCLUSIONS: The proposed framework simultaneously considers correction for confounding variables, selection of effective confounders, random effects from paired design and integration by meta-analysis. The approach improved disease-related biomarker and pathway detection, which greatly enhanced understanding of MDD neurobiology. The statistical framework can be applied to similar experimental design encountered in other complex and heterogeneous diseases.

MOTIVATION: Microarray experiments frequently produce multiple missing values (MVs) due to flaws such as dust, scratches, insufficient resolution or hybridization errors on the chips. Unfortunately, many downstream algorithms require a complete data matrix. The motivation of this work is to determine the impact of MV imputation on downstream analysis, and whether ranking of imputation methods by imputation accuracy correlates well with the biological impact of the imputation. METHODS: Using eight datasets for differential expression (DE) and classification analysis and eight datasets for gene clustering, we demonstrate the biological impact of missing-value imputation on statistical downstream analyses, including three commonly employed DE methods, four classifiers and three gene-clustering methods. Correlation between the rankings of imputation methods based on three root-mean squared error (RMSE) measures and the rankings based on the downstream analysis methods was used to investigate which RMSE measure was most consistent with the biological impact measures, and which downstream analysis methods were the most sensitive to the choice of imputation procedure. RESULTS: DE was the most sensitive to the choice of imputation procedure, while classification was the least sensitive and clustering was intermediate between the two. The logged RMSE (LRMSE) measure had the highest correlation with the imputation rankings based on the DE results, indicating that the LRMSE is the best representative surrogate among the three RMSE-based measures. Bayesian principal component analysis and least squares adaptive appeared to be the best performing methods in the empirical downstream evaluation.

MOTIVATION: Systematic information integration of multiple-related microarray studies has become an important issue as the technology becomes mature and prevalent in the past decade. The aggregated information provides more robust and accurate biomarker detection. So far, published meta-analysis methods for this purpose mostly consider two-class comparison. Methods for combining multi-class studies and considering expression pattern concordance are rarely explored. RESULTS: In this article, we develop three integration methods for biomarker detection in multiple multi-class microarray studies: ANOVA-maxP, min-MCC and OW-min-MCC. We first consider a natural extension of combining P-values from the traditional ANOVA model. Since P-values from ANOVA do not guarantee to reflect the concordant expression pattern information across studies, we propose a multi-class correlation (MCC) measure to specifically seek for biomarkers of concordant inter-class patterns across a pair of studies. For both ANOVA and MCC approaches, we use extreme order statistics to identify biomarkers differentially expressed (DE) in all studies (i.e. ANOVA-maxP and min-MCC). The min-MCC method is further extended to identify biomarkers DE in partial studies by incorporating a recently developed optimally weighted (OW) technique (OW-min-MCC). All methods are evaluated by simulation studies and by three meta-analysis applications to multi-tissue mouse metabolism datasets, multi-condition mouse trauma datasets and multi-malignant-condition human prostate cancer datasets. The results show complementary strength of the three methods for different biological purposes. AVAILABILITY: http://www.biostat.pitt.edu/bioinfo/. SUPPLEMENTARY INFORMATION: Supplementary data is available at Bioinformatics online.

MOTIVATION: Traditional genomic prediction models based on individual genes suffer from low reproducibility across microarray studies due to the lack of robustness to expression measurement noise and gene missingness when they are matched across platforms. It is common that some of the genes in the prediction model established in a training study cannot be matched to another test study because a different platform is applied. The failure of inter-study predictions has severely hindered the clinical applications of microarray. To overcome the drawbacks of traditional gene-based prediction (GBP) models, we propose a module-based prediction (MBP) strategy via unsupervised gene clustering. RESULTS: K-means clustering is used to group genes sharing similar expression profiles into gene modules, and small modules are merged into their nearest neighbors. Conventional univariate or multivariate feature selection procedure is applied and a representative gene from each selected module is identified to construct the final prediction model. As a result, the prediction model is portable to any test study as long as partial genes in each module exist in the test study. We demonstrate that K-means cluster sizes generally follow a multinomial distribution and the failure probability of inter-study prediction due to missing genes is diminished by merging small clusters into their nearest neighbors. By simulation and applications of real datasets in inter-study predictions, we show that the proposed MBP provides slightly improved accuracy while is considerably more robust than traditional GBP. AVAILABILITY: http://www.biostat.pitt.edu/bioinfo/ CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: Many pathway analysis (or gene set enrichment analysis) methods have been developed to identify enriched pathways under different biological states within a genomic study. As more and more microarray datasets accumulate, meta-analysis methods have also been developed to integrate information among multiple studies. Currently, most meta-analysis methods for combining genomic studies focus on biomarker detection and meta-analysis for pathway analysis has not been systematically pursued. RESULTS: We investigated two approaches of meta-analysis for pathway enrichment (MAPE) by combining statistical significance across studies at the gene level (MAPE_G) or at the pathway level (MAPE_P). Simulation results showed increased statistical power of meta-analysis approaches compared to a single study analysis and showed complementary advantages of MAPE_G and MAPE_P under different scenarios. We also developed an integrated method (MAPE_I) that incorporates advantages of both approaches. Comprehensive simulations and applications to real data on drug response of breast cancer cell lines and lung cancer tissues were evaluated to compare the performance of three MAPE variations. MAPE_P has the advantage of not requiring gene matching across studies. When MAPE_G and MAPE_P show complementary advantages, the hybrid version of MAPE_I is generally recommended. AVAILABILITY: http://www.biostat.pitt.edu/bioinfo/ CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: Reproducibility analyses of biologically relevant microarray studies have mostly focused on overlap of detected biomarkers or correlation of differential expression evidences across studies. For clinical utility, direct inter-study prediction (i.e. to establish a prediction model in one study and apply to another) for disease diagnosis or prognosis prediction is more important. Normalization plays a key role for such a task. Traditionally, sample-wise normalization has been a standard for inter-array and inter-study normalization. For gene-wise normalization, it has been implemented for intra-study or inter-study predictions in a few papers while its rationale, strategy and effect remain unexplored. RESULTS: In this article, we investigate the effect of gene-wise normalization in microarray inter-study prediction. Gene-specific intensity discrepancies across studies are commonly found even after proper sample-wise normalization. We explore the rationale and necessity of gene-wise normalization. We also show that the ratio of sample sizes in normal versus diseased groups can greatly affect the performance of gene-wise normalization and an analytical method is developed to adjust for the imbalanced ratio effect. Both simulation results and applications to three lung cancer and two prostate cancer data sets, considering both binary classification and survival risk predictions, showed significant and robust improvement of the new adjustment. A calibration scheme is developed to apply the ratio-adjusted gene-wise normalization for prospective clinical trials. The number of calibration samples needed is estimated from existing studies and suggested for future applications. The result has important implication to the translational research of microarray as a practical disease diagnosis and prognosis prediction tool.

MOTIVATION: Most genotyping technologies for single nucleotide polymorphism (SNP) markers use standard clustering methods to 'call' the SNP genotypes. These methods are not always optimal in distinguishing the genotype clusters of a SNP because they do not take advantage of specific features of the genotype calling problem. In particular, when family data are available, pedigree information is ignored. Furthermore, prior information about the distribution of the measurements for each cluster can be used to choose an appropriate model-based clustering method and can significantly improve the genotype calls. One special genotyping problem that has never been discussed in the literature is that of genotyping of trisomic individuals, such as individuals with Down syndrome. Calling trisomic genotypes is a more complicated problem, and the addition of external information becomes very important. RESULTS: In this article, we discuss the impact of incorporating external information into clustering algorithms to call the genotypes for both disomic and trisomic data. We also propose two new methods to call genotypes using family data. One is a modification of the K-means method and uses the pedigree information by updating all members of a family together. The other is a likelihood-based method that combines the Gaussian or beta-mixture model with pedigree information. We compare the performance of these two methods and some other existing methods using simulation studies. We also compare the performance of these methods on a real dataset generated by the Illumina platform (www.illumina.com). AVAILABILITY: The R code for the family-based genotype calling methods (SNPCaller) is available to be downloaded from the following website: http://watson.hgen.pitt.edu/register.

BACKGROUND: Gene expression data frequently contain missing values, however, most down-stream analyses for microarray experiments require complete data. In the literature many methods have been proposed to estimate missing values via information of the correlation patterns within the gene expression matrix. Each method has its own advantages, but the specific conditions for which each method is preferred remains largely unclear. In this report we describe an extensive evaluation of eight current imputation methods on multiple types of microarray experiments, including time series, multiple exposures, and multiple exposures x time series data. We then introduce two complementary selection schemes for determining the most appropriate imputation method for any given data set. RESULTS: We found that the optimal imputation algorithms (LSA, LLS, and BPCA) are all highly competitive with each other, and that no method is uniformly superior in all the data sets we examined. The success of each method can also depend on the underlying "complexity" of the expression data, where we take complexity to indicate the difficulty in mapping the gene expression matrix to a lower-dimensional subspace. We developed an entropy measure to quantify the complexity of expression matrixes and found that, by incorporating this information, the entropy-based selection (EBS) scheme is useful for selecting an appropriate imputation algorithm. We further propose a simulation-based self-training selection (STS) scheme. This technique has been used previously for microarray data imputation, but for different purposes. The scheme selects the optimal or near-optimal method with high accuracy but at an increased computational cost. CONCLUSION: Our findings provide insight into the problem of which imputation method is optimal for a given data set. Three top-performing methods (LSA, LLS and BPCA) are competitive with each other. Global-based imputation methods (PLS, SVD, BPCA) performed better on mcroarray data with lower complexity, while neighbour-based methods (KNN, OLS, LSA, LLS) performed better in data with higher complexity. We also found that the EBS and STS schemes serve as complementary and effective tools for selecting the optimal imputation algorithm.

BACKGROUND: The Signal-to-Noise-Ratio (SNR) is often used for identification of biomarkers for two-class problems and no formal and useful generalization of SNR is available for multiclass problems. We propose innovative generalizations of SNR for multiclass cancer discrimination through introduction of two indices, Gene Dominant Index and Gene Dormant Index (GDIs). These two indices lead to the concepts of dominant and dormant genes with biological significance. We use these indices to develop methodologies for discovery of dominant and dormant biomarkers with interesting biological significance. The dominancy and dormancy of the identified biomarkers and their excellent discriminating power are also demonstrated pictorially using the scatterplot of individual gene and 2-D Sammon's projection of the selected set of genes. Using information from the literature we have shown that the GDI based method can identify dominant and dormant genes that play significant roles in cancer biology. These biomarkers are also used to design diagnostic prediction systems. RESULTS AND DISCUSSION: To evaluate the effectiveness of the GDIs, we have used four multiclass cancer data sets (Small Round Blue Cell Tumors, Leukemia, Central Nervous System Tumors, and Lung Cancer). For each data set we demonstrate that the new indices can find biologically meaningful genes that can act as biomarkers. We then use six machine learning tools, Nearest Neighbor Classifier (NNC), Nearest Mean Classifier (NMC), Support Vector Machine (SVM) classifier with linear kernel, and SVM classifier with Gaussian kernel, where both SVMs are used in conjunction with one-vs-all (OVA) and one-vs-one (OVO) strategies. We found GDIs to be very effective in identifying biomarkers with strong class specific signatures. With all six tools and for all data sets we could achieve better or comparable prediction accuracies usually with fewer marker genes than results reported in the literature using the same computational protocols. The dominant genes are usually easy to find while good dormant genes may not always be available as dormant genes require stronger constraints to be satisfied; but when they are available, they can be used for authentication of diagnosis. CONCLUSION: Since GDI based schemes can find a small set of dominant/dormant biomarkers that is adequate to design diagnostic prediction systems, it opens up the possibility of using real-time qPCR assays or antibody based methods such as ELISA for an easy and low cost diagnosis of diseases. The dominant and dormant genes found by GDIs can be used in different ways to design more reliable diagnostic prediction systems.

MOTIVATION: Cluster analysis is one of the most important data mining tools for investigating high-throughput biological data. The existence of many scattered objects that should not be clustered has been found to hinder performance of most traditional clustering algorithms in such a high-dimensional complex situation. Very often, additional prior knowledge from databases or previous experiments is also available in the analysis. Excluding scattered objects and incorporating existing prior information are desirable to enhance the clustering performance. RESULTS: In this article, a class of loss functions is proposed for cluster analysis and applied in high-throughput genomic and proteomic data. Two major extensions from K-means are involved: penalization and weighting. The additive penalty term is used to allow a set of scattered objects without being clustered. Weights are introduced to account for prior information of preferred or prohibited cluster patterns to be identified. Their relationship with the classification likelihood of Gaussian mixture models is explored. Incorporation of good prior information is also shown to improve the global optimization issue in clustering. Applications of the proposed method on simulated data as well as high-throughput data sets from tandem mass spectrometry (MS/MS) and microarray experiments are presented. Our results demonstrate its superior performance over most existing methods and its computational simplicity and extensibility in the application of large complex biological data sets. AVAILABILITY: http://www.pitt.edu/~ctseng/research/software.html. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: Microarray technology has been widely applied in biological and clinical studies for simultaneous monitoring of gene expression in thousands of genes. Gene clustering analysis is found useful for discovering groups of correlated genes potentially co-regulated or associated to the disease or conditions under investigation. Many clustering methods including hierarchical clustering, K-means, PAM, SOM, mixture model-based clustering and tight clustering have been widely used in the literature. Yet no comprehensive comparative study has been performed to evaluate the effectiveness of these methods. RESULTS: In this paper, six gene clustering methods are evaluated by simulated data from a hierarchical log-normal model with various degrees of perturbation as well as four real datasets. A weighted Rand index is proposed for measuring similarity of two clustering results with possible scattered genes (i.e. a set of noise genes not being clustered). Performance of the methods in the real data is assessed by a predictive accuracy analysis through verified gene annotations. Our results show that tight clustering and model-based clustering consistently outperform other clustering methods both in simulated and real data while hierarchical clustering and SOM perform among the worst. Our analysis provides deep insight to the complicated gene clustering problem of expression profile and serves as a practical guideline for routine microarray cluster analysis.

BACKGROUND: The quality of microarray data can seriously affect the accuracy of downstream analyses. In order to reduce variability and enhance signal reproducibility in these data, many normalization methods have been proposed and evaluated, most of which are for data obtained from cDNA microarrays and Affymetrix GeneChips. CodeLink Bioarrays are a newly emerged, single-color oligonucleotide microarray platform. To date, there are no reported studies that evaluate normalization methods for CodeLink Bioarrays. RESULTS: We compared five existing normalization approaches, in terms of both noise reduction and signal retention: Median (suggested by the manufacturer), CyclicLoess, Quantile, Iset, and Qspline. These methods were applied to two real datasets (a time course dataset and a lung disease-related dataset) generated by CodeLink Bioarrays and were assessed using multiple statistical significance tests. Compared to Median, CyclicLoess and Qspline exhibit a significant and the most consistent improvement in reduction of variability and retention of signal. CyclicLoess appears to retain more signal than Qspline. Quantile reduces more variability than Median in both datasets, yet fails to consistently retain more signal in the time course dataset. Iset does not improve over Median in either noise reduction or signal enhancement in the time course dataset. CONCLUSION: Median is insufficient either to reduce variability or to retain signal effectively for CodeLink Bioarray data. CyclicLoess is a more suitable approach for normalizing these data. CyclicLoess also seems to be the most effective method among the five different normalization strategies examined.