High-dimensional Gene Expression Data Research Articles

A two-stage sparse logistic regression for optimal gene selection in high-dimensional microarray data classification

The common issues of high-dimensional gene expression data are that many of the genes may not be relevant, and there exists a high correlation among genes. Gene selection has been proven to be an effective way to improve the results of many classification methods. Sparse logistic regression using least absolute shrinkage and selection operator (lasso) or using smoothly clipped absolute deviation is one of the most widely applicable methods in cancer classification for gene selection. However, this method faces a critical challenge in practical applications when there are high correlations among genes. To address this problem, a two-stage sparse logistic regression is proposed, with the aim of obtaining an efficient subset of genes with high classification capabilities by combining the screening approach as a filter method and adaptive lasso with a new weight as an embedded method. In the first stage, sure independence screening method as a screening approach retains those genes representing high individual correlation with the cancer class level. In the second stage, the adaptive lasso with new weight is implemented to address the existence of high correlations among the screened genes in the first stage. Experimental results based on four publicly available gene expression datasets have shown that the proposed method significantly outperforms three state-of-the-art methods in terms of classification accuracy, G-mean, area under the curve, and stability. In addition, the results demonstrate that the top selected genes are biologically related to the cancer type. Thus, the proposed method can be useful for cancer classification using DNA gene expression data in real clinical practice.

Robust clustering of noisy high-dimensional gene expression data for patients subtyping.

One of the most important research areas in personalized medicine is the discovery of disease sub-types with relevance in clinical applications. This is usually accomplished by exploring gene expression data with unsupervised clustering methodologies. Then, with the advent of multiple omics technologies, data integration methodologies have been further developed to obtain better performances in patient separability. However, these methods do not guarantee the survival separability of the patients in different clusters. We propose a new methodology that first computes a robust and sparse correlation matrix of the genes, then decomposes it and projects the patient data onto the first m spectral components of the correlation matrix. After that, a robust and adaptive to noise clustering algorithm is applied. The clustering is set up to optimize the separation between survival curves estimated cluster-wise. The method is able to identify clusters that have different omics signatures and also statistically significant differences in survival time. The proposed methodology is tested on five cancer datasets downloaded from The Cancer Genome Atlas repository. The proposed method is compared with the Similarity Network Fusion (SNF) approach, and model based clustering based on Student's t-distribution (TMIX). Our method obtains a better performance in terms of survival separability, even if it uses a single gene expression view compared to the multi-view approach of the SNF method. Finally, a pathway based analysis is accomplished to highlight the biological processes that differentiate the obtained patient groups. Our R source code is available online at https://github.com/angy89/RobustClusteringPatientSubtyping. Supplementary data are available at Bioinformatics online.

Gene set analysis methods: a systematic comparison

BackgroundGene set analysis is a valuable tool to summarize high-dimensional gene expression data in terms of biologically relevant sets. This is an active area of research and numerous gene set analysis methods have been developed. Despite this popularity, systematic comparative studies have been limited in scope.MethodsIn this study we present a semi-synthetic simulation study using real datasets in order to test and compare commonly used methods.ResultsA software pipeline, Flexible Algorithm for Novel Gene set Simulation (FANGS) develops simulated data based on a prostate cancer dataset where the KRAS and TGF-β pathways were differentially expressed. The FANGS software is compatible with other datasets and pathways. Comparisons of gene set analysis methods are presented for Gene Set Enrichment Analysis (GSEA), Significance Analysis of Function and Expression (SAFE), sigPathway, and Correlation Adjusted Mean RAnk (CAMERA) methods. All gene set analysis methods are tested using gene sets from the MSigDB knowledge base. The false positive rate and power are estimated and presented for comparison. Recommendations are made for the utility of the default settings of methods and each method’s sensitivity towards various effect sizes.ConclusionsThe results of this study provide empirical guidance to users of gene set analysis methods. The FANGS software is available for researchers for continued methods comparisons.

A big data pipeline: Identifying dynamic gene regulatory networks from time-course Gene Expression Omnibus data with applications to influenza infection.

A biological host response to an external stimulus or intervention such as a disease or infection is a dynamic process, which is regulated by an intricate network of many genes and their products. Understanding the dynamics of this gene regulatory network allows us to infer the mechanisms involved in a host response to an external stimulus, and hence aids the discovery of biomarkers of phenotype and biological function. In this article, we propose a modeling/analysis pipeline for dynamic gene expression data, called Pipeline4DGEData, which consists of a series of statistical modeling techniques to construct dynamic gene regulatory networks from the large volumes of high-dimensional time-course gene expression data that are freely available in the Gene Expression Omnibus repository. This pipeline has a consistent and scalable structure that allows it to simultaneously analyze a large number of time-course gene expression data sets, and then integrate the results across different studies. We apply the proposed pipeline to influenza infection data from nine studies and demonstrate that interesting biological findings can be discovered with its implementation.

Gene Selection Using High Dimensional Gene Expression Data: An Appraisal

Gene Selection Using High Dimensional Gene Expression Data: An Appraisal

Edge-group sparse PCA for network-guided high dimensional data analysis.

Principal component analysis (PCA) has been widely used to deal with high-dimensional gene expression data. In this study, we proposed an Edge-group Sparse PCA (ESPCA) model by incorporating the group structure from a prior gene network into the PCA framework for dimension reduction and feature interpretation. ESPCA enforces sparsity of principal component (PC) loadings through considering the connectivity of gene variables in the prior network. We developed an alternating iterative algorithm to solve ESPCA. The key of this algorithm is to solve a new k-edge sparse projection problem and a greedy strategy has been adapted to address it. Here we adopted ESPCA for analyzing multiple gene expression matrices simultaneously. By incorporating prior knowledge, our method can overcome the drawbacks of sparse PCA and capture some gene modules with better biological interpretations. We evaluated the performance of ESPCA using a set of artificial datasets and two real biological datasets (including TCGA pan-cancer expression data and ENCODE expression data), and compared their performance with PCA and sparse PCA. The results showed that ESPCA could identify more biologically relevant genes, improve their biological interpretations and reveal distinct sample characteristics. An R package of ESPCA is available at http://page.amss.ac.cn/shihua.zhang/. Supplementary data are available at Bioinformatics online.

Application of Penalized Mixed Model in Identification of Genes in Yeast Cell-Cycle Gene Expression Datas

High-dimensional time-course gene expression data refer to time course data with a large number of covariates. In this status, variable selection is a popular approach for selecting important variables. In this paper, we review penalized likelihood mixed effects model for variable selection in high-dimensional time-course data. Then, the approach is used for variable selection in yeast cell-cycle gene expression data

Statistical approach for selection of biologically informative genes

Selection of informative genes from high dimensional gene expression data has emerged as an important research area in genomics. Many gene selection techniques have been proposed so far are either based on relevancy or redundancy measure. Further, the performance of these techniques has been adjudged through post selection classification accuracy computed through a classifier using the selected genes. This performance metric may be statistically sound but may not be biologically relevant. A statistical approach, i.e. Boot-MRMR, was proposed based on a composite measure of maximum relevance and minimum redundancy, which is both statistically sound and biologically relevant for informative gene selection. For comparative evaluation of the proposed approach, we developed two biological sufficient criteria, i.e. Gene Set Enrichment with QTL (GSEQ) and biological similarity score based on Gene Ontology (GO). Further, a systematic and rigorous evaluation of the proposed technique with 12 existing gene selection techniques was carried out using five gene expression datasets. This evaluation was based on a broad spectrum of statistically sound (e.g. subject classification) and biological relevant (based on QTL and GO) criteria under a multiple criteria decision-making framework. The performance analysis showed that the proposed technique selects informative genes which are more biologically relevant. The proposed technique is also found to be quite competitive with the existing techniques with respect to subject classification and computational time. Our results also showed that under the multiple criteria decision-making setup, the proposed technique is best for informative gene selection over the available alternatives. Based on the proposed approach, an R Package, i.e. BootMRMR has been developed and available at https://cran.r-project.org/web/packages/BootMRMR. This study will provide a practical guide to select statistical techniques for selecting informative genes from high dimensional expression data for breeding and system biology studies.

Survey on gene selection methodologies based on filter, wrapper, embedded and hybrid approaches

Gene selection has become a most elementary tool for processing high dimensional data. It is an important pre-processing step which is essential for microarray analysis and classification. DNA microarray technique is applied for analyzing large number of genes at a time, by which the expression levels of the genes are determined. Selection of genes applying high dimensional gene expression data is primary and most important work for accurate disease prediction and classification. Gene expression data are in the form of matrix and usually it contains irrelevant, redundant and noisy data, so the study and analysis of data faces problematic situation. Most important and primary purpose of gene selection methods is to get rid of from the problem of curse of dimensionality, enhance the performance and efficiency of classification algorithms by eliminating irrelevant genes and reduction of noise. This paper describes the various gene selection methods declaring with their relevant pros and cons. It also denotes a review on few gene selection methods, mainly those that have been proposed in last few years.

Integrating transcriptional activity in genome-scale models of metabolism

BackgroundGenome-scale metabolic models provide an opportunity for rational approaches to studies of the different reactions taking place inside the cell. The integration of these models with gene regulatory networks is a hot topic in systems biology. The methods developed to date focus mostly on resolving the metabolic elements and use fairly straightforward approaches to assess the impact of genome expression on the metabolic phenotype.ResultsWe present here a method for integrating the reverse engineering of gene regulatory networks into these metabolic models. We applied our method to a high-dimensional gene expression data set to infer a background gene regulatory network. We then compared the resulting phenotype simulations with those obtained by other relevant methods.ConclusionsOur method outperformed the other approaches tested and was more robust to noise. We also illustrate the utility of this method for studies of a complex biological phenomenon, the diauxic shift in yeast.

Mixture Models With a Prior on the Number of Components

ABSTRACTA natural Bayesian approach for mixture models with an unknown number of components is to take the usual finite mixture model with symmetric Dirichlet weights, and put a prior on the number of components—that is, to use a mixture of finite mixtures (MFM). The most commonly used method of inference for MFMs is reversible jump Markov chain Monte Carlo, but it can be nontrivial to design good reversible jump moves, especially in high-dimensional spaces. Meanwhile, there are samplers for Dirichlet process mixture (DPM) models that are relatively simple and are easily adapted to new applications. It turns out that, in fact, many of the essential properties of DPMs are also exhibited by MFMs—an exchangeable partition distribution, restaurant process, random measure representation, and stick-breaking representation—and crucially, the MFM analogues are simple enough that they can be used much like the corresponding DPM properties. Consequently, many of the powerful methods developed for inference in DPMs can be directly applied to MFMs as well; this simplifies the implementation of MFMs and can substantially improve mixing. We illustrate with real and simulated data, including high-dimensional gene expression data used to discriminate cancer subtypes. Supplementary materials for this article are available online.

A Self-Training Subspace Clustering Algorithm under Low-Rank Representation for Cancer Classification on Gene Expression Data.

Accurate identification of the cancer types is essential to cancer diagnoses and treatments. Since cancer tissue and normal tissue have different gene expression, gene expression data can be used as an efficient feature source for cancer classification. However, accurate cancer classification directly using original gene expression profiles remains challenging due to the intrinsic high-dimension feature and the small size of the data samples. We proposed a new self-training subspace clustering algorithm under low-rank representation, called SSC-LRR, for cancer classification on gene expression data. Low-rank representation (LRR) is first applied to extract discriminative features from the high-dimensional gene expression data; the self-training subspace clustering (SSC) method is then used to generate the cancer classification predictions. The SSC-LRR was tested on two separate benchmark datasets in control with four state-of-the-art classification methods. It generated cancer classification predictions with an overall accuracy 89.7 percent and a general correlation 0.920, which are 18.9 and 24.4 percent higher than that of the best control method respectively. In addition, several genes (RNF114, HLA-DRB5, USP9Y, and PTPN20) were identified by SSC-LRR as new cancer identifiers that deserve further clinical investigation. Overall, the study demonstrated a new sensitive avenue to recognize cancer classifications from large-scale gene expression data.

Subspace Weighting Co-Clustering of Gene Expression Data.

Microarray technology enables the collection of vast amounts of gene expression data from biological experiments. Clustering algorithms have been successfully applied to exploring the gene expression data. Since a set of genes may be only correlated to a subset of samples, it is useful to use co-clustering to recover co-clusters in the gene expression data. In this paper, we propose a novel algorithm, called Subspace Weighting Co-Clustering (SWCC), for high dimensional gene expression data. In SWCC, a gene subspace weight matrix is introduced to identify the contribution of gene objects in distinguishing different sample clusters. We design a new co-clustering objective function to recover the co-clusters in the gene expression data, in which the subspace weight matrix is introduced. An iterative algorithm is developed to solve the objective function, in which the subspace weight matrix is automatically computed during the iterative co-clustering process. Our empirical study shows encouraging results of the proposed algorithm in comparison with six state-of-the-art clustering algorithms on ten gene expression data sets. We also propose to use SWCC for gene clustering and selection. The experimental results show that the selected genes can improve the classification performance of Random Forests.

Studies on the Clustering Algorithm for Analyzing Gene Expression Data with a Bidirectional Penalty.

This article reports a new clustering method based on the k-means algorithm to high-dimensional gene expression data. The proposed approach makes use of bidirectional penalties to constrain the number of clusters and centroids of clusters to simultaneously determine the unknown number of clusters and handle large amounts of noise in gene expression data. Numeric studies indicate that this algorithm not only performs better in clustering but is also comparable to other approaches in its ability to obtain the correct number of clusters and correct signal features. Finally, we apply the proposed approach to analyze two benchmark gene expression datasets. These analyses again indicate that the proposed algorithm performs well in clustering high-dimensional gene expression data with an unknown number of clusters.

Learning Parsimonious Classification Rules from Gene Expression Data Using Bayesian Networks with Local Structure.

The comprehensibility of good predictive models learned from high-dimensional gene expression data is attractive because it can lead to biomarker discovery. Several good classifiers provide comparable predictive performance but differ in their abilities to summarize the observed data. We extend a Bayesian Rule Learning (BRL-GSS) algorithm, previously shown to be a significantly better predictor than other classical approaches in this domain. It searches a space of Bayesian networks using a decision tree representation of its parameters with global constraints, and infers a set of IF-THEN rules. The number of parameters and therefore the number of rules are combinatorial to the number of predictor variables in the model. We relax these global constraints to a more generalizable local structure (BRL-LSS). BRL-LSS entails more parsimonious set of rules because it does not have to generate all combinatorial rules. The search space of local structures is much richer than the space of global structures. We design the BRL-LSS with the same worst-case time-complexity as BRL-GSS while exploring a richer and more complex model space. We measure predictive performance using Area Under the ROC curve (AUC) and Accuracy. We measure model parsimony performance by noting the average number of rules and variables needed to describe the observed data. We evaluate the predictive and parsimony performance of BRL-GSS, BRL-LSS and the state-of-the-art C4.5 decision tree algorithm, across 10-fold cross-validation using ten microarray gene-expression diagnostic datasets. In these experiments, we observe that BRL-LSS is similar to BRL-GSS in terms of predictive performance, while generating a much more parsimonious set of rules to explain the same observed data. BRL-LSS also needs fewer variables than C4.5 to explain the data with similar predictive performance. We also conduct a feasibility study to demonstrate the general applicability of our BRL methods on the newer RNA sequencing gene-expression data.

Statistical Approaches for Gene Selection, Hub Gene Identification and Module Interaction in Gene Co-Expression Network Analysis: An Application to Aluminum Stress in Soybean (Glycine max L.).

Selection of informative genes is an important problem in gene expression studies. The small sample size and the large number of genes in gene expression data make the selection process complex. Further, the selected informative genes may act as a vital input for gene co-expression network analysis. Moreover, the identification of hub genes and module interactions in gene co-expression networks is yet to be fully explored. This paper presents a statistically sound gene selection technique based on support vector machine algorithm for selecting informative genes from high dimensional gene expression data. Also, an attempt has been made to develop a statistical approach for identification of hub genes in the gene co-expression network. Besides, a differential hub gene analysis approach has also been developed to group the identified hub genes into various groups based on their gene connectivity in a case vs. control study. Based on this proposed approach, an R package, i.e., dhga (https://cran.r-project.org/web/packages/dhga) has been developed. The comparative performance of the proposed gene selection technique as well as hub gene identification approach was evaluated on three different crop microarray datasets. The proposed gene selection technique outperformed most of the existing techniques for selecting robust set of informative genes. Based on the proposed hub gene identification approach, a few number of hub genes were identified as compared to the existing approach, which is in accordance with the principle of scale free property of real networks. In this study, some key genes along with their Arabidopsis orthologs has been reported, which can be used for Aluminum toxic stress response engineering in soybean. The functional analysis of various selected key genes revealed the underlying molecular mechanisms of Aluminum toxic stress response in soybean.

Rank-based classifiers for extremely high-dimensional gene expression data

Predicting phenotypes on the basis of gene expression profiles is a classification task that is becoming increasingly important in the field of precision medicine. Although these expression signals are real-valued, it is questionable if they can be analyzed on an interval scale. As with many biological signals their influence on e.g. protein levels is usually non-linear and thus can be misinterpreted. In this article we study gene expression profiles with up to 54,000 dimensions. We analyze these measurements on an ordinal scale by replacing the real-valued profiles by their ranks. This type of rank transformation can be used for the construction of invariant classifiers that are not affected by noise induced by data transformations which can occur in the measurement setup. Our 10 $$\times $$ 10 fold cross-validation experiments on 86 different data sets and 19 different classification models indicate that classifiers largely benefit from this transformation. Especially random forests and support vector machines achieve improved classification results on a significant majority of datasets.

Improving sensitivity of linear regression-based cell type-specific differential expression deconvolution with per-gene vs. global significance threshold

BackgroundThe goal of many human disease-oriented studies is to detect molecular mechanisms different between healthy controls and patients. Yet, commonly used gene expression measurements from blood samples suffer from variability of cell composition. This variability hinders the detection of differentially expressed genes and is often ignored. Combined with cell counts, heterogeneous gene expression may provide deeper insights into the gene expression differences on the cell type-specific level.Published computational methods use linear regression to estimate cell type-specific differential expression, and a global cutoff to judge significance, such as False Discovery Rate (FDR). Yet, they do not consider many artifacts hidden in high-dimensional gene expression data that may negatively affect linear regression. In this paper we quantify the parameter space affecting the performance of linear regression (sensitivity of cell type-specific differential expression detection) on a per-gene basis.ResultsWe evaluated the effect of sample sizes, cell type-specific proportion variability, and mean squared error on sensitivity of cell type-specific differential expression detection using linear regression. Each parameter affected variability of cell type-specific expression estimates and, subsequently, the sensitivity of differential expression detection. We provide the R package, LRCDE, which performs linear regression-based cell type-specific differential expression (deconvolution) detection on a gene-by-gene basis. Accounting for variability around cell type-specific gene expression estimates, it computes per-gene t-statistics of differential detection, p-values, t-statistic-based sensitivity, group-specific mean squared error, and several gene-specific diagnostic metrics.ConclusionsThe sensitivity of linear regression-based cell type-specific differential expression detection differed for each gene as a function of mean squared error, per group sample sizes, and variability of the proportions of target cell (cell type being analyzed). We demonstrate that LRCDE, which uses Welch’s t-test to compare per-gene cell type-specific gene expression estimates, is more sensitive in detecting cell type-specific differential expression at α < 0.05 missed by the global false discovery rate threshold FDR < 0.3.Electronic supplementary materialThe online version of this article (doi:10.1186/s12859-016-1226-z) contains supplementary material, which is available to authorized users.

Estimation of random effects and identifying heterogeneous genes in meta-analysis of gene expression studies.

Combining effect sizes from individual studies using random-effects meta-analysis models are commonly applied in high-dimensional gene expression data. However, unknown study heterogeneity can arise from inconsistencies in sample quality and experimental conditions. High heterogeneity of effect sizes can reduce statistical power of the models. In this study, we describe three hypothesis-testing frameworks for meta-analysis of microarray data, and review several existing meta-analytic techniques that have been used in the genomic setting. These include P-value-based methods, rank-based methods and effect-size-based methods. We then discuss limitations of some of these methods and describe random-effects-based methods in detail. We introduce two methods for estimating the inter-study variance in random-effects meta-analytic models and another method for identifying heterogeneous genes for gene expression data. We compared various methods with the standard and existing meta-analytic techniques in the genomic framework. We demonstrate our results through a series of simulations and application in Alzheimer's gene expression data.

Traveling on discrete embeddings of gene expression.

High-throughput technologies have generated an unprecedented amount of high-dimensional gene expression data. Algorithmic approaches could be extremely useful to distill information and derive compact interpretable representations of the statistical patterns present in the data. This paper proposes a mining approach to extract an informative representation of gene expression profiles based on a generative model called the Counting Grid (CG). Using the CG model, gene expression values are arranged on a discrete grid, learned in a way that "similar" co-expression patterns are arranged in close proximity, thus resulting in an intuitive visualization of the dataset. More than this, the model permits to identify the genes that distinguish between classes (e.g. different types of cancer). Finally, each sample can be characterized with a discriminative signature - extracted from the model - that can be effectively employed for classification. A thorough evaluation on several gene expression datasets demonstrate the suitability of the proposed approach from a twofold perspective: numerically, we reached state-of-the-art classification accuracies on 5 datasets out of 7, and similar results when the approach is tested in a gene selection setting (with a stability always above 0.87); clinically, by confirming that many of the genes highlighted by the model as significant play also a key role for cancer biology. The proposed framework can be successfully exploited to meaningfully visualize the samples; detect medically relevant genes; properly classify samples.