PCA disjoint models for multiclass cancer analysis using gene expression data

MOTIVATION: Microarray expression profiling appears particularly promising for a deeper understanding of cancer biology and to identify molecular signatures supporting the histological classification schemes of neoplastic specimens. However, molecular diagnostics based on microarray data presents major challenges due to the overwhelming number of variables and the complex, multiclass nature of tumor samples. Thus, the development of marker selection methods, that allow the identification of those genes that are most likely to confer high classification accuracy of multiple tumor types, and of multiclass classification schemes is of paramount importance. RESULTS: A computational procedure for marker identification and for classification of multiclass gene expression data through the application of disjoint principal component models is described. The identified features represent a rational and dimensionally reduced base for understanding the basic biology of diseases, defining targets for therapeutic intervention, and developing diagnostic tools for the identification and classification of multiple pathological states. The method has been tested on different microarray data sets obtained from various human tumor samples. The results demonstrate that this procedure allows the identification of specific phenotype markers and can classify previously unseen instances in the presence of multiple classes.     

Prediction of biologically significant components from microarray data: Independently Consistent Expression Discriminator (ICED)

MOTIVATION: Class distinction is a supervised learning approach that has been successfully employed in the analysis of high-throughput gene expression data. Identification of a set of genes that predicts differential biological states allows for the development of basic and clinical scientific approaches to the diagnosis of disease. The Independent Consistent Expression Discriminator (ICED) was designed to provide a more biologically relevant search criterion during predictor selection by embracing the inherent variability of gene expression in any biological state. The four components of ICED include (i) normalization of raw data; (ii) assignment of weights to genes from both classes; (iii) counting of votes to determine optimal number of predictor genes for class distinction; (iv) calculation of prediction strengths for classification results. The search criteria employed by ICED is designed to identify not only genes that are consistently expressed at one level in one class and at a consistently different level in another class but identify genes that are variable in one class and consistent in another. The result is a novel approach to accurately select biologically relevant predictors of differential disease states from a small number of microarray samples. RESULTS: The data described herein utilized ICED to analyze the large AML/ALL training and test data set (Golub et al., 1999, Science, 286, 531-537) in addition to a smaller data set consisting of an animal model of the childhood neurodegenerative disorder, Batten disease, generated for this study. Both of the analyses presented herein have correctly predicted biologically relevant perturbations that can be used for disease classification, irrespective of sample size. Furthermore, the results have provided candidate proteins for future study in understanding the disease process and the identification of potential targets for therapeutic intervention.     

Neural progenitor genes. Germinal zone expression and analysis of genetic overlap in stem cell populations

The identification of the genes regulating neural progenitor cell (NPC) functions is of great importance to developmental neuroscience and neural repair. Previously, we combined genetic subtraction and microarray analysis to identify genes enriched in neural progenitor cultures. Here, we apply a strategy to further stratify the neural progenitor genes. In situ hybridization demonstrates expression in the central nervous system germinal zones of 54 clones so identified, making them highly relevant for study in brain and neural progenitor development. Using microarray analysis we find 73 genes enriched in three neural stem cell (NSC)-containing populations generated under different conditions. We use the custom microarray to identify 38 "stemness" genes, with enriched expression in the three NSC conditions and present in both embryonic stem cells and hematopoietic stem cells. However, comparison of expression profiles from these stem cell populations indicates that while there is shared gene expression, the amount of genetic overlap is no more than what would be expected by chance, indicating that different stem cells have largely different gene expression patterns. Taken together, these studies identify many genes not previously associated with neural progenitor cell biology and also provide a rational scheme for stratification of microarray data for functional analysis.     

Classification of multiple cancer types by multicategory support vector machines using gene expression data

MOTIVATION: High-density DNA microarray measures the activities of several thousand genes simultaneously and the gene expression profiles have been used for the cancer classification recently. This new approach promises to give better therapeutic measurements to cancer patients by diagnosing cancer types with improved accuracy. The Support Vector Machine (SVM) is one of the classification methods successfully applied to the cancer diagnosis problems. However, its optimal extension to more than two classes was not obvious, which might impose limitations in its application to multiple tumor types. We briefly introduce the Multicategory SVM, which is a recently proposed extension of the binary SVM, and apply it to multiclass cancer diagnosis problems. RESULTS: Its applicability is demonstrated on the leukemia data (Golub et al., 1999) and the small round blue cell tumors of childhood data (Khan et al., 2001). Comparable classification accuracy shown in the applications and its flexibility render the MSVM a viable alternative to other classification methods. SUPPLEMENTARY INFORMATION: http://www.stat.ohio-state.edu/~yklee/msvm.htm     

Differential gene expression analysis by DNA microarrays technology and its application in molecular oncology

Accumulation of genetic and epigenetic aberrations leads to malignant transformation of normal cells. Functional studies of cancer using genomic and proteomic tools will help to reveal the true complexity of the processes leading to cancer development in humans. Until recently, diagnosis and prognosis of cancer was based on conventional pathologic criteria and epidemiological evidence. Certain tumors were divided only into relatively broad histological and morphological subcategories. Rapidly developing methods of differential gene expression analysis promote the search for clinically relevant genes changing their expression levels during malignant transformation. DNA microarrays offer a unique possibility to rapidly assess the global expression picture of thousands genes in any given time point and compare the detailed combinatory analysis results of global expression profiles for normal and malignant cells at various functional stages or separate experimental conditions. Acquisition of such "genetic portraits" allows searching for regularity and difference in expression patterns of certain genes, understanding their function and pathological importance, and ultimately developing the "molecular nosology" of cancer. This review describes the basis of DNA microarray technology and methodology, and focuses on their applications in molecular classification of tumors, drug sensitivity and resistance studies, and identification of biological markers of cancer.     

Differential gene expression detection and sample classification using penalized linear regression models

Differential gene expression detection and sample classification using microarray data have received much research interest recently. Owing to the large number of genes p and small number of samples n (p > n), microarray data analysis poses big challenges for statistical analysis. An obvious problem owing to the 'large p small n' is over-fitting. Just by chance, we are likely to find some non-differentially expressed genes that can classify the samples very well. The idea of shrinkage is to regularize the model parameters to reduce the effects of noise and produce reliable inferences. Shrinkage has been successfully applied in the microarray data analysis. The SAM statistics proposed by Tusher et al. and the 'nearest shrunken centroid' proposed by Tibshirani et al. are ad hoc shrinkage methods. Both methods are simple, intuitive and prove to be useful in empirical studies. Recently Wu proposed the penalized t/F-statistics with shrinkage by formally using the (1) penalized linear regression models for two-class microarray data, showing good performance. In this paper we systematically discussed the use of penalized regression models for analyzing microarray data. We generalize the two-class penalized t/F-statistics proposed by Wu to multi-class microarray data. We formally derive the ad hoc shrunken centroid used by Tibshirani et al. using the (1) penalized regression models. And we show that the penalized linear regression models provide a rigorous and unified statistical framework for sample classification and differential gene expression detection.     

Tissue classification with gene expression profiles.

Constantly improving gene expression profiling technologies are expected to provide understanding and insight into cancer-related cellular processes. Gene expression data is also expected to significantly aid in the development of efficient cancer diagnosis and classification platforms. In this work we examine three sets of gene expression data measured across sets of tumor(s) and normal clinical samples: The first set consists of 2,000 genes, measured in 62 epithelial colon samples (Alon et al., 1999). The second consists of approximately equal to 100,000 clones, measured in 32 ovarian samples (unpublished extension of data set described in Schummer et al. (1999)). The third set consists of approximately equal to 7,100 genes, measured in 72 bone marrow and peripheral blood samples (Golub et al, 1999). We examine the use of scoring methods, measuring separation of tissue type (e.g., tumors from normals) using individual gene expression levels. These are then coupled with high-dimensional classification methods to assess the classification power of complete expression profiles. We present results of performing leave-one-out cross validation (LOOCV) experiments on the three data sets, employing nearest neighbor classifier, SVM (Cortes and Vapnik, 1995), AdaBoost (Freund and Schapire, 1997) and a novel clustering-based classification technique. As tumor samples can differ from normal samples in their cell-type composition, we also perform LOOCV experiments using appropriately modified sets of genes, attempting to eliminate the resulting bias. We demonstrate success rate of at least 90% in tumor versus normal classification, using sets of selected genes, with, as well as without, cellular-contamination-related members. These results are insensitive to the exact selection mechanism, over a certain range.     

Deriving quantitative conclusions from microarray expression data

MOTIVATION: The last few years have seen the development of DNA microarray technology that allows simultaneous measurement of the expression levels of thousands of genes. While many methods have been developed to analyze such data, most have been visualization-based. Methods that yield quantitative conclusions have been diverse and complex. RESULTS: We present two straightforward methods for identifying specific genes whose expression is linked with a phenotype or outcome variable as well as for systematically predicting sample class membership: (1) a conservative, permutation-based approach to identifying differentially expressed genes; (2) an augmentation of K-nearest-neighbor pattern classification. Our analyses replicate the quantitative conclusions of Golub et al. (1999; Science, 286, 531-537) on leukemia data, with better classification results, using far simpler methods. With the breast tumor data of Perou et al. (2000; Nature, 406, 747-752), the methods lend rigorous quantitative support to the conclusions of the original paper. In the case of the lymphoma data in Alizadeh et al. (2000; Nature, 403, 503-511), our analyses only partially support the conclusions of the original authors. AVAILABILITY: The software and supplementary information are available freely to researchers at academic and non-profit institutions at http://cc.ucsf.edu/jain/public     

Mixture modelling of gene expression data from microarray experiments

MOTIVATION: Hierarchical clustering is one of the major analytical tools for gene expression data from microarray experiments. A major problem in the interpretation of the output from these procedures is assessing the reliability of the clustering results. We address this issue by developing a mixture model-based approach for the analysis of microarray data. Within this framework, we present novel algorithms for clustering genes and samples. One of the byproducts of our method is a probabilistic measure for the number of true clusters in the data. RESULTS: The proposed methods are illustrated by application to microarray datasets from two cancer studies; one in which malignant melanoma is profiled (Bittner et al., Nature, 406, 536-540, 2000), and the other in which prostate cancer is profiled (Dhanasekaran et al., 2001, submitted).     

Gene discovery analysis from mouse embryonic stem cells based on time course microarray data

An embryonic stem cell is a powerful tool for investigation of early development in vitro. The study of embryonic stem cell mediated neuronal differentiation allows for improved understanding of the mechanisms involved in embryonic neuronal development. We investigated expression profile changes using time course cDNA microarray to identify clues for the signaling network of neuronal differentiation. For the short time course microarray data, pattern analysis based on the quadratic regression method is an effective approach for identification and classification of a variety of expressed genes that have biological relevance. We studied the expression patterns, at each of 5 stages, after neuronal induction at the mRNA level of embryonic stem cells using the quadratic regression method for pattern analysis. As a result, a total of 316 genes (3.1%) including 166 (1.7%) informative genes in 8 possible expression patterns were identified by pattern analysis. Among the selected genes associated with neurological system, all three genes showing linearly increasing pattern over time, and one gene showing decreasing pattern over time, were verified by RT-PCR. Therefore, an increase in gene expression over time, in a linear pattern, may be associated with embryonic development. The genes: Tcfap2c, Ttr, Wnt3a, Btg2 and Foxk1 detected by pattern analysis, and verified by RT-PCR simultaneously, may be candidate markers associated with the development of the nervous system. Our study shows that pattern analysis, using the quadratic regression method, is very useful for investigation of time course cDNA microarray data. The pattern analysis used in this study has biological significance for the study of embryonic stem cells.     

Identifying new targets in leukemogenesis using computational approaches

There is a need to identify novel targets in Acute Lymphoblastic Leukemia (ALL), a hematopoietic cancer affecting children, to improve our understanding of disease biology and that can be used for developing new therapeutics. Hence, the aim of our study was to find new genes as targets using in silico studies; for this we retrieved the top 10% overexpressed genes from Oncomine public domain microarray expression database; 530 overexpressed genes were short-listed from Oncomine database. Then, using prioritization tools such as ENDEAVOUR, DIR and TOPPGene online tools, we found fifty-four genes common to the three prioritization tools which formed our candidate leukemogenic genes for this study. As per the protocol we selected thirty training genes from PubMed. The prioritized and training genes were then used to construct STRING functional association network, which was further analyzed using cytoHubba hub analysis tool to investigate new genes which could form drug targets in leukemia. Analysis of the STRING protein network built from these prioritized and training genes led to identification of two hub genes, SMAD2 and CDK9, which were not implicated in leukemogenesis earlier. Filtering out from several hundred genes in the network we also found MEN1, HDAC1 and LCK genes, which re-emphasized the important role of these genes in leukemogenesis. This is the first report on these five additional signature genes in leukemogenesis. We propose these as new targets for developing novel therapeutics and also as biomarkers in leukemogenesis, which could be important for prognosis and diagnosis.     

Investigation of differentially expressed genes during the development of mouse cerebellum

Before the discovery of DNA microarray and DNA chip technology, the expression of only a small number of genes could be analyzed at a time. Currently, such technology allows us the simultaneous analysis of a large number of genes to systematically monitor their expression patterns that may be associated with various biological phenomena. We utilized the Affymetrix GeneChip Mu11K to analyze the gene expression profile in developing mouse cerebellum to assist in the understanding of the genetic basis of cerebellar development in mice. Our analysis showed 81.6% (10,321/12,654) of the genes represented on the GeneChip were expressed in the postnatal cerebellum, and among those, 8.7% (897/10,321) were differentially expressed with more than a two-fold change in their maximum and minimum expression levels during the developmental time course. Further analysis of the differentially expressed genes that were clustered in terms of their expression patterns and the function of their encoded products revealed an aspect of the genetic foundation that lies beneath the cellular events and neural network formation that takes place during the development of the mouse cerebellum.     

Detecting differentially expressed genes by relative entropy

DNA microarray experiments have generated large amount of gene expression measurements across different conditions. One crucial step in the analysis of these data is to detect differentially expressed genes. Some parametric methods, including the two-sample t-test (T-test) and variations of it, have been used. Alternatively, a class of non-parametric algorithms, such as the Wilcoxon rank sum test (WRST), significance analysis of microarrays (SAM) of Tusher et al. (2001), the empirical Bayesian (EB) method of Efron et al. (2001), etc., have been proposed. Most available popular methods are based on t-statistic. Due to the quality of the statistic that they used to describe the difference between groups of data, there are situations when these methods are inefficient, especially when the data follows multi-modal distributions. For example, some genes may display different expression patterns in the same cell type, say, tumor or normal, to form some subtypes. Most available methods are likely to miss these genes. We developed a new non-parametric method for selecting differentially expressed genes by relative entropy, called SDEGRE, to detect differentially expressed genes by combining relative entropy and kernel density estimation, which can detect all types of differences between two groups of samples. The significance of whether a gene is differentially expressed or not can be estimated by resampling-based permutations. We illustrate our method on two data sets from Golub et al. (1999) and Alon et al. (1999). Comparing the results with those of the T-test, the WRST and the SAM, we identified novel differentially expressed genes which are of biological significance through previous biological studies while they were not detected by the other three methods. The results also show that the genes selected by SDEGRE have a better capability to distinguish the two cell types.     

Inference of a genetic network by a combined approach of cluster analysis and graphical Gaussian modeling

MOTIVATION: Recent advances in DNA microarray technologies have made it possible to measure the expression levels of thousands of genes simultaneously under different conditions. The data obtained by microarray analyses are called expression profile data. One type of important information underlying the expression profile data is the 'genetic network,' that is, the regulatory network among genes. Graphical Gaussian Modeling (GGM) is a widely utilized method to infer or test relationships among a plural of variables. RESULTS: In this study, we developed a method combining the cluster analysis with GGM for the inference of the genetic network from the expression profile data. The expression profile data of 2467 Saccharomyces cerevisiae genes measured under 79 different conditions (Eisen et al., PROC: Natl Acad. Sci. USA, 95, 14683-14868, 1998) were used for this study. At first, the 2467 genes were classified into 34 clusters by a cluster analysis, as a preprocessing for GGM. Then, the expression levels of the genes in each cluster were averaged for each condition. The averaged expression profile data of 34 clusters were subjected to GGM, and a partial correlation coefficient matrix was obtained as a model of the genetic network of S. cerevisiae. The accuracy of the inferred network was examined by the agreement of our results with the cumulative results of experimental studies.     

Drug Repositioning through Systematic Mining of Gene Coexpression Networks in Cancer

Gene coexpression network analysis is a powerful “data-driven” approach essential for understanding cancer biology and mechanisms of tumor development. Yet, despite the completion of thousands of studies on cancer gene expression, there have been few attempts to normalize and integrate co-expression data from scattered sources in a concise “meta-analysis” framework. We generated such a resource by exploring gene coexpression networks in 82 microarray datasets from 9 major human cancer types. The analysis was conducted using an elaborate weighted gene coexpression network (WGCNA) methodology and identified over 3,000 robust gene coexpression modules. The modules covered a range of known tumor features, such as proliferation, extracellular matrix remodeling, hypoxia, inflammation, angiogenesis, tumor differentiation programs, specific signaling pathways, genomic alterations, and biomarkers of individual tumor subtypes. To prioritize genes with respect to those tumor features, we ranked genes within each module by connectivity, leading to identification of module-specific functionally prominent hub genes. To showcase the utility of this network information, we positioned known cancer drug targets within the coexpression networks and predicted that Anakinra, an anti-rheumatoid therapeutic agent, may be promising for development in colorectal cancer. We offer a comprehensive, normalized and well documented collection of >3000 gene coexpression modules in a variety of cancers as a rich data resource to facilitate further progress in cancer research.     

Differential Gene Expression Analysis by DNA Microarray Technology and Its Application in Molecular Oncology

Accumulation of genetic and epigenetic aberrations leads to malignant transformation of normal cells. Functional studies of cancer using genomic and proteomic tools aim to reveal the true complexity of the processes leading to cancer development in humans. Until recently, diagnosis and prognosis of cancer was based on conventional pathologic criteria and epidemiological evidence. Certain tumors were divided only into relatively broad histological and morphological subcategories. Rapidly developing methods of differential gene expression analysis promote the search for clinically relevant genes changing their expression levels during malignant transformation. DNA microarrays offer a unique possibility to rapidly assess the global expression picture of thousands genes in any given time point and compare the results of detailed combinatory analysis of global expression profiles for normal and malignant cells at various functional stages or separate experimental conditions. Acquisition of such genetic portraits allows searching for regularity and difference in expression patterns of certain genes, understanding their function and pathological importance, and ultimately developing the molecular nosology of cancer. This review describes the basis of DNA microarray technology and methodology, and focuses on their application in molecular classification of tumors, drug sensitivity and resistance studies, and identification of biological markers of cancer.