Significance of gene ranking for classification of microarray samples

Many methods for classification and gene selection with microarray data have been developed. These methods usually give a ranking of genes. Evaluating the statistical significance of the gene ranking is important for understanding the results and for further biological investigations, but this question has not been well addressed for machine learning methods in existing works. Here, we address this problem by formulating it in the framework of hypothesis testing and propose a solution based on resampling. The proposed r-test methods convert gene ranking results into position p-values to evaluate the significance of genes. The methods are tested on three real microarray data sets and three simulation data sets with support vector machines as the method of classification and gene selection. The obtained position p-values help to determine the number of genes to be selected and enable scientists to analyze selection results by sophisticated multivariate methods under the same statistical inference paradigm as for simple hypothesis testing methods.     

On gene ranking using replicated microarray time course data

Summary .  Consider the ranking of genes using data from replicated microarray time course experiments, where there are multiple biological conditions, and the genes of interest are those whose temporal profiles differ across conditions. We derive a multisample multivariate empirical Bayes' statistic for ranking genes in the order of differential expression, from both longitudinal and cross-sectional replicated developmental microarray time course data. Our longitudinal multisample model assumes that time course replicates are independent and identically distributed multivariate normal vectors. On the other hand, we construct a cross-sectional model using a normal regression framework with any appropriate basis for the design matrices. In both cases, we use natural conjugate priors in our empirical Bayes' setting which guarantee closed form solutions for the posterior odds. The simulations and two case studies using published worm and mouse microarray time course datasets indicate that the proposed approaches perform satisfactorily.     

Classification across gene expression microarray studies

Background  

The increasing number of gene expression microarray studies represents an important resource in biomedical research. As a result, gene expression based diagnosis has entered clinical practice for patient stratification in breast cancer. However, the integration and combined analysis of microarray studies remains still a challenge. We assessed the potential benefit of data integration on the classification accuracy and systematically evaluated the generalization performance of selected methods on four breast cancer studies comprising almost 1000 independent samples. To this end, we introduced an evaluation framework which aims to establish good statistical practice and a graphical way to monitor differences. The classification goal was to correctly predict estrogen receptor status (negative/positive) and histological grade (low/high) of each tumor sample in an independent study which was not used for the training. For the classification we chose support vector machines (SVM), predictive analysis of microarrays (PAM), random forest (RF) and k-top scoring pairs (kTSP). Guided by considerations relevant for classification across studies we developed a generalization of kTSP which we evaluated in addition. Our derived version (DV) aims to improve the robustness of the intrinsic invariance of kTSP with respect to technologies and preprocessing.     

Assessing stability of gene selection in microarray data analysis

Background  

The number of genes declared differentially expressed is a random variable and its variability can be assessed by resampling techniques. Another important stability indicator is the frequency with which a given gene is selected across subsamples. We have conducted studies to assess stability and some other properties of several gene selection procedures with biological and simulated data.     

Entropy-based gene ranking without selection bias for the predictive classification of microarray data

Background

We describe the E-RFE method for gene ranking, which is useful for the identification of markers in the predictive classification of array data. The method supports a practical modeling scheme designed to avoid the construction of classification rules based on the selection of too small gene subsets (an effect known as the selection bias, in which the estimated predictive errors are too optimistic due to testing on samples already considered in the feature selection process).

Results

With E-RFE, we speed up the recursive feature elimination (RFE) with SVM classifiers by eliminating chunks of uninteresting genes using an entropy measure of the SVM weights distribution. An optimal subset of genes is selected according to a two-strata model evaluation procedure: modeling is replicated by an external stratified-partition resampling scheme, and, within each run, an internal K-fold cross-validation is used for E-RFE ranking. Also, the optimal number of genes can be estimated according to the saturation of Zipf's law profiles.

Conclusions

Without a decrease of classification accuracy, E-RFE allows a speed-up factor of 100 with respect to standard RFE, while improving on alternative parametric RFE reduction strategies. Thus, a process for gene selection and error estimation is made practical, ensuring control of the selection bias, and providing additional diagnostic indicators of gene importance.

     

Reproducibility-optimized test statistic for ranking genes in microarray studies

A principal goal of microarray studies is to identify the genes showing differential expression under distinct conditions. In such studies, the selection of an optimal test statistic is a crucial challenge, which depends on the type and amount of data under analysis. While previous studies on simulated or spike-in datasets do not provide practical guidance on how to choose the best method for a given real dataset, we introduce an enhanced reproducibility-optimization procedure, which enables the selection of a suitable gene- anking statistic directly from the data. In comparison with existing ranking methods, the reproducibilityoptimized statistic shows good performance consistently under various simulated conditions and on Affymetrix spike-in dataset. Further, the feasibility of the novel statistic is confirmed in a practical research setting using data from an in-house cDNA microarray study of asthma-related gene expression changes. These results suggest that the procedure facilitates the selection of an appropriate test statistic for a given dataset without relying on a priori assumptions, which may bias the findings and their interpretation. Moreover, the general reproducibilityoptimization procedure is not limited to detecting differential expression only but could be extended to a wide range of other applications as well.     

An investigation of biomarkers derived from legacy microarray data for their utility in the RNA-seq era

Background

Gene expression microarray has been the primary biomarker platform ubiquitously applied in biomedical research, resulting in enormous data, predictive models, and biomarkers accrued. Recently, RNA-seq has looked likely to replace microarrays, but there will be a period where both technologies co-exist. This raises two important questions: Can microarray-based models and biomarkers be directly applied to RNA-seq data? Can future RNA-seq-based predictive models and biomarkers be applied to microarray data to leverage past investment?

Results

We systematically evaluated the transferability of predictive models and signature genes between microarray and RNA-seq using two large clinical data sets. The complexity of cross-platform sequence correspondence was considered in the analysis and examined using three human and two rat data sets, and three levels of mapping complexity were revealed. Three algorithms representing different modeling complexity were applied to the three levels of mappings for each of the eight binary endpoints and Cox regression was used to model survival times with expression data. In total, 240,096 predictive models were examined.

Conclusions

Signature genes of predictive models are reciprocally transferable between microarray and RNA-seq data for model development, and microarray-based models can accurately predict RNA-seq-profiled samples; while RNA-seq-based models are less accurate in predicting microarray-profiled samples and are affected both by the choice of modeling algorithm and the gene mapping complexity. The results suggest continued usefulness of legacy microarray data and established microarray biomarkers and predictive models in the forthcoming RNA-seq era.

Electronic supplementary material

The online version of this article (doi:10.1186/s13059-014-0523-y) contains supplementary material, which is available to authorized users.     

Context-specific infinite mixtures for clustering gene expression profiles across diverse microarray dataset

MOTIVATION: Identifying groups of co-regulated genes by monitoring their expression over various experimental conditions is complicated by the fact that such co-regulation is condition-specific. Ignoring the context-specific nature of co-regulation significantly reduces the ability of clustering procedures to detect co-expressed genes due to additional 'noise' introduced by non-informative measurements. RESULTS: We have developed a novel Bayesian hierarchical model and corresponding computational algorithms for clustering gene expression profiles across diverse experimental conditions and studies that accounts for context-specificity of gene expression patterns. The model is based on the Bayesian infinite mixtures framework and does not require a priori specification of the number of clusters. We demonstrate that explicit modeling of context-specificity results in increased accuracy of the cluster analysis by examining the specificity and sensitivity of clusters in microarray data. We also demonstrate that probabilities of co-expression derived from the posterior distribution of clusterings are valid estimates of statistical significance of created clusters. AVAILABILITY: The open-source package gimm is available at http://eh3.uc.edu/gimm.     

The microarray analysis for gene expression in haploids and diploids derived from twin-seedling rice

In this study, microarray technique was employed to analyze the gene expression at the RNA level between haploids and corresponding diploids derived from a rice twin-seedling line SARII-628. Differ- ent degrees of expression variations were observed in the plant after haploidization. The main results are as follows: (1) after haploidization, the ratio of the sensitive loci was 2.47% of the total loci designed on chip. Those loci were randomly distributed on the 12 pairs of rice chromosomes and the activated loci were more than the silenced ones. (2) Gene clusters on chromosome were observed for 33 se- quences. (3) GoPipe function classification for 575 sensitive loci revealed an involvement in the bio- logical process, cell component and molecular function. (4) RT-PCR generally validated the result from microarray with a coincidence rate of 83.78%. And for the randomly-selected activated or silenced loci in chip analysis, the coincidence rate was up to 91.86%.     

Adjusting confounders in ranking biomarkers: a model-based ROC approach

High-throughput studies have been extensively conducted in the research of complex human diseases. As a representative example, consider gene-expression studies where thousands of genes are profiled at the same time. An important objective of such studies is to rank the diagnostic accuracy of biomarkers (e.g. gene expressions) for predicting outcome variables while properly adjusting for confounding effects from low-dimensional clinical risk factors and environmental exposures. Existing approaches are often fully based on parametric or semi-parametric models and target evaluating estimation significance as opposed to diagnostic accuracy. Receiver operating characteristic (ROC) approaches can be employed to tackle this problem. However, existing ROC ranking methods focus on biomarkers only and ignore effects of confounders. In this article, we propose a model-based approach which ranks the diagnostic accuracy of biomarkers using ROC measures with a proper adjustment of confounding effects. To this end, three different methods for constructing the underlying regression models are investigated. Simulation study shows that the proposed methods can accurately identify biomarkers with additional diagnostic power beyond confounders. Analysis of two cancer gene-expression studies demonstrates that adjusting for confounders can lead to substantially different rankings of genes.     

Sample size calculations based on ranking and selection in microarray experiments

Summary .   We develop formulae to calculate sample sizes for ranking and selection of differentially expressed genes among different clinical subtypes or prognostic classes of disease in genome-wide screening studies with microarrays. The formulae aim to control the probability that a selected subset of genes with fixed size contains enough truly top-ranking informative genes, which can be assessed on the basis of the distribution of ordered statistics from independent genes. We provide strategies for conservative designs to cope with issues of unknown number of informative genes and unknown correlation structure across genes. Application of the formulae to a clinical study for multiple myeloma is given.     

The microarray analysis for gene expression in haploids and diploids derived from twin-seedling rice

In this study, microarray technique was employed to analyze the gene expression at the RNA level between haploids and corresponding diploids derived from a rice twin-seedling line SARII-628. Different degrees of expression variations were observed in the plant after haploidization. The main results are as follows: (1) after haploidization, the ratio of the sensitive loci was 2.47% of the total loci designed on chip. Those loci were randomly distributed on the 12 pairs of rice chromosomes and the activated loci were more than the silenced ones. (2) Gene clusters on chromosome were observed for 33 sequences. (3) GoPipe function classification for 575 sensitive loci revealed an involvement in the biological process, cell component and molecular function. (4) RT-PCR generally validated the result from microarray with a coincidence rate of 83.78%. And for the randomly-selected activated or silenced loci in chip analysis, the coincidence rate was up to 91.86%.     

Extended rat liver slice survival and stability monitored using clinical biomarkers

Precision-cut liver slices are reportedly limited as toxicity models by their short half-life in culture. We used traditional clinical chemistry biomarkers and histology to assess a newer procedure for improved liver slice maintenance. Slices from Sprague-Dawley rat livers were well maintained in a roller culture system for up to 10 days based on protein content (60-70% or higher of initial values) and biomarker retention and verified by histological examination of the tissues showing morphological integrity and viability of hepatocyte and biliary regions. Exposure of the slices to geldanamycin (GEL) resulted in loss of slice LDH and transaminase content, with associated depression in ALP and GGT levels and elevated bilirubin, indicating that GEL affects both cell types as occurs in vivo with this hepatobiliary toxicant. Thus, we conclude that liver slices merit further investigation as a general model for chronic as well as acute toxicity studies.     

Reliable gene signatures for microarray classification: assessment of stability and performance

MOTIVATION: Two important questions for the analysis of gene expression measurements from different sample classes are (1) how to classify samples and (2) how to identify meaningful gene signatures (ranked gene lists) exhibiting the differences between classes and sample subsets. Solutions to both questions have immediate biological and biomedical applications. To achieve optimal classification performance, a suitable combination of classifier and gene selection method needs to be specifically selected for a given dataset. The selected gene signatures can be unstable and the resulting classification accuracy unreliable, particularly when considering different subsets of samples. Both unstable gene signatures and overestimated classification accuracy can impair biological conclusions. METHODS: We address these two issues by repeatedly evaluating the classification performance of all models, i.e. pairwise combinations of various gene selection and classification methods, for random subsets of arrays (sampling). A model score is used to select the most appropriate model for the given dataset. Consensus gene signatures are constructed by extracting those genes frequently selected over many samplings. Sampling additionally permits measurement of the stability of the classification performance for each model, which serves as a measure of model reliability. RESULTS: We analyzed a large gene expression dataset with 78 measurements of four different cartilage sample classes. Classifiers trained on subsets of measurements frequently produce models with highly variable performance. Our approach provides reliable classification performance estimates via sampling. In addition to reliable classification performance, we determined stable consensus signatures (i.e. gene lists) for sample classes. Manual literature screening showed that these genes are highly relevant to our gene expression experiment with osteoarthritic cartilage. We compared our approach to others based on a publicly available dataset on breast cancer. AVAILABILITY: R package at http://www.bio.ifi.lmu.de/~davis/edaprakt     

CGO: utilizing and integrating gene expression microarray data in clinical research and data management

Clinical GeneOrganizer (CGO) is a novel windows-based archiving, organization and data mining software for the integration of gene expression profiling in clinical medicine. The program implements various user-friendly tools and extracts data for further statistical analysis. This software was written for Affymetrix GeneChip *.txt files, but can also be used for any other microarray-derived data. The MS-SQL server version acts as a data mart and links microarray data with clinical parameters of any other existing database and therefore represents a valuable tool for combining gene expression analysis and clinical disease characteristics.     

Cluster stability scores for microarray data in cancer studies

Background  

A potential benefit of profiling of tissue samples using microarrays is the generation of molecular fingerprints that will define subtypes of disease. Hierarchical clustering has been the primary analytical tool used to define disease subtypes from microarray experiments in cancer settings. Assessing cluster reliability poses a major complication in analyzing output from clustering procedures. While most work has focused on estimating the number of clusters in a dataset, the question of stability of individual-level clusters has not been addressed.     

Quantifying difference in gene expression profile between bovine blastocysts derived by in vitro fertilization and somatic cell nuclear transfer

Epigenetic reprogramming intensely occurs in somatic-cell nuclear transfer (SCNT) embryos, which highlights the importance of proper expressions of reprogramming-related genes in SCNT embryos. We here assessed gene expression profiles (GEPs) difference between bovine blastocyst groups derived by in-vitro fertilization (IVF) or SCNT; in SCNT, cumulus cells and ear skin fibroblasts were used for cSCNT and fSCNT blastocysts, respectively. We obtained GEPs of 15 reprogramming-related genes in single blastocysts using multiplex PCR and found a broad range of variations in their GEPs. Weighted root-mean-square deviation (wRMSD) analysis, which calculates the deviation of SCNT blastocysts' GEPs from IVF blastocysts' mean GEP, found a significant difference between IVF and fSCNT and between cSCNT and fSCNT blastocysts (p < 0.001) but not between IVF and cSCNT. Since the fibroblasts' GEP was more distant from the IVF blastocysts' than the cumulus cells', it might partly explain the less similarity of fSCNT blastocysts' GEPs to the IVF's mean GEP. Our wRMSD method succeeds in expressing in figures how different two comparable embryo groups of different derivations are in GEP, which would be useful to select a better embryo derivation protocol among the candidates prior to field applications.