Invited Keynote Talk: Data Mining and Statistical Methods for Analyzing Microarray Experiments

A reliable and precise classification of tumors is essential for successful diagnosis and treatment of cancer. cDNA microarrays and high-density oligonucleotide chips are novel biotechnologies increasingly used in cancer research. By allowing the monitoring of expression levels in cells for thousands of genes simultaneously, microarray experiments may lead to a more complete understanding of the molecular variations among tumors and hence to a finer and more informative classification. The ability to successfully distinguish between tumor classes (already known or yet to be discovered) using gene expression data is an important aspect of this novel approach to cancer classification. This article compares the performance of different discrimination methods for the classification of tumors based on gene expression data. The methods include nearest-neighbor classifiers, linear discriminant analysis, and classification trees. Recent machine learning approaches, such as bagging and boosting, are also considered. The discrimination methods are applied to datasets from three recently published cancer gene expression studies.

With the recent development of biotechnologies, cDNA microarray chips are increasingly applied in cancer research. Microarray experiments can lead to a more thorough grasp of the molecular variations among tumors because they can allow the monitoring of expression levels in cells for thousands of genes simultaneously. Accordingly, how to successfully discriminate the classes of tumors using gene expression data is an urgent research issue and plays an important role in carcinogenesis. To refine the large dimension of the genes data and effectively classify tumor classes, this study proposes several hybrid discrimination procedures that combine the statistical-based techniques and computational intelligence approaches to discriminate the tumor classes. A real microarray data set was used to demonstrate the performance of the proposed approaches. In addition, the results of cross-validation experiments reveal that the proposed two-stage hybrid models are more efficient in discriminating the acute leukemia classes than the established single stage models.

Deoxyribonucleic acid (DNA) microarrays are collections of DNA probes arranged on a base pair and the latest commercialized molecular diagnostic technologies that offer high throughput results, more sensitive and require less time. It is the most reliable and widely accepted tool facilitating the simultaneous identification of thousands of genetic elements even a single gene. Microarrays are powerful new tools for the investigation of global changes in gene expression profiles in cells and tissues. The different types of DNA microarray or DNA chip devices and systems are described along with their methods of fabrication and their use. The DNA microarrays assembly process is automatized and further miniaturized. DNA microarrays are used in the search of various specific genes or in gene polymorphism and expression analysis. They will be widely used to investigate the expression of various genes connected with various diseases in order to find the causes of these diseases and to enable their accurate treatment. Generally, microarray analysis is not only applied for gene expression studies, but also used in immunology, genotyping, diagnostics and sequence analysis. Additionally, microarray technology being developed and applied to new areas of proteomics, cancer research, and cellular analysis.

Gene expression data generated from microarray experiments are characterized by large number of genes or dimensions. Informative gene selection for performing clustering to discover useful phenotypes is a major issue as there is no class information available. In this paper, we propose a wrapper based feature selection approach to perform sample based clustering on gene expression data. The proposed work uses Particle Swarm Optimization(PSO) for best subset generation and k-means as wrapper algorithm for evaluating the subsets. Experimental results show that the features selected by this method is able to produce clusters of good quality. Clustering accuracy of 70–80% were obtained for different datasets.

The advent of the so-called cDNA microarrays has offered the first possibility to obtain a global understanding of biological processes in living organisms by simultaneous readouts of tens of thousands of genes. Initial experiments suggest that genes with similar function have similar expression patterns in microarray experiments. Until now, most approaches to computational analysis of gene expressions have used unsupervised learning. Although in some cases unsupervised methods may be suficient, the complexity of the biological processes is so high that it is unlikely that purely syntactical analyses are capable of fully exploiting the richness of the microarray data. In addition, it seems natural to re-use the existing biological (background) knowledge. In this paper, we present some elements of a methodology for knowledge discovery from microarray experiments. Two source of bio-medical knowledge are used: Ashburner’s gene ontology and our own literature-derived network of gene-gene relations obtained by analysing Medline citation records. Predictive models can be induced and their classification quality validated through the ROC/AUC analysis and applied to provide hypotheses regarding the function of unclassified genes. The methodology has been so far tested on publicly available gene expression data and its results evaluated by molecular biologists and medical researchers.KeywordsKnowledge DiscoverySqualene EpoxidaseComputational Molecular BiologyNorwegian Radium HospitalUniversal AcademyThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Microarray databases are the most frequently used datasets for cancer analytics. Microarray databases are characterized by the presence of a very large number of genes, which exceeds the very little number of samples. So, the feature set accumulates the curse of dimensionality. Therefore, selecting a small subset of genes among thousands of genes in microarray data can potentially increase the accuracy for the classification of cancer. Many approaches, from the field of classical machine learning and soft computing, have been used to address the issue of feature selection and feature extraction for better classifications and clustering accuracy. The research outlined in this paper strives to look at a two-stage approach using minimum Redundancy Maximum Relevancy (mRMR), a feature ranking framework as the first stage followed by a hybrid genetic algorithm in the second stage that works on the features ranked by the mRMR. The proposed method is aimed to select the optimal feature subsets for better classification results in binary and multi class datasets to compensate for the curse of dimensionality in microarray datasets. The classifiers used to test the two-stage proposition are SVM, Naive-Bayes, Linear Discriminant Analysis, decision trees and random forest classifiers. The experimental results show that the gene subset selected by the mRMR-GA pipeline gives good results.

The classification of tissue samples based on gene expression data is an important problem in medical diagnosis of diseases such as cancer. In gene expression data, the number of genes is usually very high (in the thousands) compared to the number of data samples (in the tens or low hundreds); that is, the data dimension is large compared to the number of data points (such data is said to be undersampled). To cope with performance and accuracy problems associated with high dimensionality, it is commonplace to apply a preprocessing step that transforms the data to a space of significantly lower dimension with limited loss of the information present in the original data. Linear Discriminant Analysis (LDA) is a well-known technique for dimension reduction and feature extraction, but it is not applicable for undersampled data due to singularity problems associated with the matrices in the underlying representation. This paper presents a dimension reduction and feature extraction scheme, called Uncorrelated Linear Discriminant Analysis (ULDA), for undersampled problems and illustrates its utility on gene expression data. ULDA employs the Generalized Singular Value Decomposition method to handle undersampled data and the features that it produces in the transformed space are uncorrelated, which makes it attractive for gene expression data. The properties of ULDA are established rigorously and extensive experimental results on gene expression data are presented to illustrate its effectiveness in classifying tissue samples. These results provide a comparative study of various state-of-the-art classification methods on well-known gene expression data sets.

Microarrays are capable of determining the expression levels of thousands of genes simultaneously. One important application of gene expression data is classification of samples into categories. In combination with classification methods, this technology can be useful to support clinical management decisions for individual patients, e.g. in oncology. Standard statistic methodologies in classification or prediction do not work well when the number of variables p (genes) far too exceeds the number of samples n. So, modification of existing statistical methodologies or development of new methodologies is needed for the analysis of microarray data. This paper proposes a new method for tumor classification using gene expression data. In this method, we first employ independent component analysis to model the gene expression data, then apply optimal scoring algorithm to classify them. Further speaking, this approach can first make full use of the high-order statistical information contained in the gene expression data. Second, this approach also employs regularized regression models to handle the situation of large numbers of correlated predictor variables. Finally, the predictive models are developed for classifying tumors based on the entire gene expression profile. To show the validity of the proposed method, we apply it to classify four DNA microarray datasets involving various human normal and tumor tissue samples. The experimental results show that the method is efficient and feasible. Matlab scripts are available on request.

Recently, expression quantitative loci (eQTL) mapping studies, where expression levels of thousands of genes are viewed as quantitative traits, have been used to provide greater insight into the biology of gene regulation. Originally, eQTLs were detected by applying standard QTL detection tools (using a "one gene at-a-time" approach), but this method ignores many possible interactions between genes. Several other methods have proposed to overcome these limitations, but each of them has some specific disadvantages. In this paper, we present an integrated hierarchical Bayesian model that jointly models all genes and SNPs to detect eQTLs. We propose a model (named iBMQ) that is specifically designed to handle a large number G of gene expressions, a large number S of regressors (genetic markers) and a small number n of individuals in what we call a ``large G, large S, small n'' paradigm. This method incorporates genotypic and gene expression data into a single model while 1) specifically coping with the high dimensionality of eQTL data (large number of genes), 2) borrowing strength from all gene expression data for the mapping procedures, and 3) controlling the number of false positives to a desirable level. To validate our model, we have performed simulation studies and showed that it outperforms other popular methods for eQTL detection, including QTLBIM, R-QTL, remMap and M-SPLS. Finally, we used our model to analyze a real expression dataset obtained in a panel of mice BXD Recombinant Inbred (RI) strains. Analysis of these data with iBMQ revealed the presence of multiple hotspots showing significant enrichment in genes belonging to one or more annotation categories.

Research in cancer genomics has proliferated with the advent of microarray technologies. These technologies facilitate monitoring of thousands of genes in parallel, thus providing insight into disease subtypes and gene functions. Gene expression data obtained from microarray chips are typified by few samples and a large number of genes. Supervised classifiers such as support vector machines (SVM) have been deployed for prediction task. However, insufficient labeled data have resulted in a paradigm shift to semi-supervised learning, in particular, transductive SVM (TSVM). Analysis of gene expression data using TSVM revealed that the performance of the model degenerates in the presence of unlabeled data. We address this issue by using a representative sampling strategy which ensures safety of the classifier even in the presence of unlabeled data. We also address the issue of privacy violation when classifier is shipped to other medical institutes for analysis of shared data. We propose a safety aware and privacy preserving TSVM for classifying cancer subtypes. Performance of TSVM with SVM and accuracy loss of the proposed TSVM are also analyzed.

ABSTRACTAmong many classification methods, linear discriminant analysis (LDA) is a favored tool due to its simplicity, robustness, and predictive accuracy but when the number of genes is larger than the number of observations, it cannot be applied directly because the within-class covariance matrix is singular. Also, diagonal LDA (DLDA) is a simpler model compared to LDA and has better performance in some cases. However, in reality, DLDA requires a strong assumption based on mutual independence. In this article, we propose the modified LDA (MLDA). MLDA is based on independence, but uses the information that has an effect on classification performance with the dependence structure. We suggest two approaches. One is the case of using gene rank. The other involves no use of gene rank. We found that MLDA has better performance than LDA, DLDA, or K-nearest neighborhood and is comparable with support vector machines in real data analysis and the simulation study.

The real-time system of the JT-60 has a cascade structure with the major and the minor control loops. The multi-variable feedback control is adopted in the major control loop in order to adjust non-linear effects which will appear in the high-/beta/region. By applying the concept of 'phase shift control' to the control program of the major control loop, a flexible tokamak discharge control is expected to be performed with taking account of machine protection.

In recent years, many studies have shown that microarray gene expression data is useful for disease identification and cancer classification. However, since gene expression data may contain thousands of genes simultaneously, successful microarray classification can be rather difficult. Feature (gene) selection is a frequently used pre-processing technology for successful classification of microarray gene expression data. Selecting a useful gene subset as a classifier not only decreases the computational time and cost, but also increases the classification accuracy. It is therefore imperative to extract only a small number of genes, which are exclusively relevant for the classification of a particular cancer/disease type. In this paper, correlation-based binary particle swarm optimizations is proposed to select the relevant genes, and a K-nearest neighbor with the leave-one-out cross-validation method serves as a classifier to evaluate the classification performance on six published cancer classification data sets. The experimental results show that the proposed method selects fewer gene subsets, while still resulting in higher prediction accuracy than the other literature methods.

Current studies document the effectiveness of multi-sensing technique implementation to mimic or to complement human senses. This work demonstrated the successful application of multi-sensing techniques such electronic tongue (e-tongue), electronic nose (e-nose) and Fourier Transform Infrared Spectroscopy (FTIR). The fusion of these modalities enhance the classification of pure Tualang honey using Linear Discriminant Analysis (LDA), Probabilistic Neural Network (PNN), Support Vector Machine (SVM) and k-Nearest Neighbour (KNN). KNN and PNN are able to classify between pure and adulterated honey samples, outperform LDA and SVM. By performing data fusion, SVM and LDA classifier can achieved more than 80% accuracy, while KNN and PNN obtained greater precision, up to 96% correct classification. The findings confirmed that, multi-sensing technique; either KNN or PNN was significantly superior compared to SVM and LDA classification methods. Thus, both analyses are able to discriminate between pure and adulterated honey.

BackgroundIn microarray gene expression profiling experiments, differentially expressed genes (DEGs) are detected from among tens of thousands of genes on an array using statistical tests. It is important to control the number of false positives or errors that are present in the resultant DEG list. To date, more than 20 different multiple test methods have been reported that compute overall Type I error rates in microarray experiments. However, these methods share the following dilemma: they have low power in cases where only a small number of DEGs exist among a large number of total genes on the array.ResultsThis study contrasts parallel multiplicity of objectively related tests against the traditional simultaneousness of subjectively related tests and proposes a new assessment called the Error Discovery Rate (EDR) for evaluating multiple test comparisons in microarray experiments. Parallel multiple tests use only the negative genes that parallel the positive genes to control the error rate; while simultaneous multiple tests use the total unchanged gene number for error estimates. Here, we demonstrate that the EDR method exhibits improved performance over other methods in specificity and sensitivity in testing expression data sets with sequence digital expression confirmation, in examining simulation data, as well as for three experimental data sets that vary in the proportion of DEGs. The EDR method overcomes a common problem of previous multiple test procedures, namely that the Type I error rate detection power is low when the total gene number used is large but the DEG number is small.ConclusionsMicroarrays are extensively used to address many research questions. However, there is potential to improve the sensitivity and specificity of microarray data analysis by developing improved multiple test comparisons. This study proposes a new view of multiplicity in microarray experiments and the EDR provides an alternative multiple test method for Type I error control in microarray experiments.