A Survey on Filter Techniques for Feature Selection in Gene Expression Microarray Analysis
IEEE/ACM Transactions on Computational Biology and BioinformaticsPublished 14 February 2012
Cosmin Lazar, Jonatan Taminau, Stijn Meganck, David Steenhoff, Alain Coletta, Colin Molter
Citations618
SJR quartileQ2
SJR score0.80
SNIP0.97
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
This survey focuses on filter feature selection methods for informative feature discovery in gene expression microarray (GEM) analysis, which is also known as differentially expressed genes (DEGs) discovery, gene prioritization, or biomarker discovery, and presents them in a unified framework.
Abstract
A plenitude of feature selection (FS) methods is available in the literature, most of them rising as a need to analyze data of very high dimension, usually hundreds or thousands of variables. Such data sets are now available in various application areas like combinatorial chemistry, text mining, multivariate imaging, or bioinformatics. As a general accepted rule, these methods are grouped in filters, wrappers, and embedded methods. More recently, a new group of methods has been added in the general framework of FS: ensemble techniques. The focus in this survey is on filter feature selection methods for informative feature discovery in gene expression microarray (GEM) analysis, which is also known as differentially expressed genes (DEGs) discovery, gene prioritization, or biomarker discovery. We present them in a unified framework, using standardized notations in order to reveal their technical details and to highlight their common characteristics as well as their particularities.
Keywords
Biochemistry, Genetics and Molecular Biology
Journal of the Royal Statistical Society Series B (Statistical Methodology)Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing
108,736 Citations1995Yoav Benjamini, Yosef Hochberg
Statistical Applications in Genetics and Molecular BiologyLinear Models and Empirical Bayes Methods for Assessing Differential Expression in Microarray Experiments
12,048 Citations2004Gordon K. Smyth
ScienceMolecular Classification of Cancer: Class Discovery and Class Prediction by Gene Expression Monitoring
11,622 Citations1999Todd R. Golub, Donna K. Slonim +10 more
A generic approach to cancer classification based on gene expression monitoring by DNA microarrays is described and applied to human acute leukemias as a test case and suggests a general strategy for discovering and predicting cancer classes for other types of cancer, independent of previous biological knowledge.
BMC Research NotesIn silico prediction of protein-protein interactions in human macrophages
10,952 Citations2014Oussema Souiai, Fatma Z. Guerfali +3 more
This work integrated the PPI dataset from the Agile Protein Interaction DataAnalyzer with different meta-data to infer a contextualized macrophage-specific interactome using a combination of statistical methods and confirms that contextualizing interactomes improves the biological significance of bioinformatic analyses.
Proceedings of the National Academy of SciencesSignificance analysis of microarrays applied to the ionizing radiation response
10,671 Citations2001Virginia Goss Tusher, Robert Tibshirani +1 more
The Annals of Mathematical StatisticsOn Estimation of a Probability Density Function and Mode
10,504 Citations1962Emanuel Parzen
Machine LearningGene Selection for Cancer Classification using Support Vector Machines
9,839 Citations2002Isabelle Guyon, Jason Weston +2 more
This paper proposes a new method of gene selection utilizing Support Vector Machine methods based on Recursive Feature Elimination (RFE), and demonstrates experimentally that the genes selected yield better classification performance and are biologically relevant to cancer.
Journal of Machine Learning ResearchAn introduction to variable and feature selection
7,831 Citations2003GuyonIsabelle, ElisseeffAndré
An introduction to variable and feature selection
6,577 Citations2009Isabelle Guyon
The relationship between Precision-Recall and ROC curves
6,255 Citations2006Jesse Davis, Mark Goadrich
It is shown that a deep connection exists between ROC space and PR space, such that a curve dominates in R OC space if and only if it dominates in PR space.
Journal of the Royal Statistical Society Series B (Statistical Methodology)A Direct Approach to False Discovery Rates
5,787 Citations2002John D. Storey
The calculation of the q‐value is discussed, the pFDR analogue of the p‐value, which eliminates the need to set the error rate beforehand as is traditionally done, and can yield an increase of over eight times in power compared with the Benjamini–Hochberg FDR method.
BioinformaticsA review of feature selection techniques in bioinformatics
5,289 Citations2007Yvan Saeys, Iñaki Inza +1 more
A basic taxonomy of feature selection techniques is provided, providing their use, variety and potential in a number of both common as well as upcoming bioinformatics applications.
ScienceExploring the Metabolic and Genetic Control of Gene Expression on a Genomic Scale
4,500 Citations1997Joseph L. DeRisi, Vishwanath R. Iyer +1 more
DNA microarrays containing virtually every gene of Saccharomyces cerevisiae were used to carry out a comprehensive investigation of the temporal program of gene expression accompanying the metabolic shift from fermentation to respiration, and the expression patterns of many previously uncharacterized genes provided clues to their possible functions.
Proceedings of the National Academy of SciencesDiagnosis of multiple cancer types by shrunken centroids of gene expression
2,947 Citations2002Robert Tibshirani, Trevor Hastie +2 more
The method of “nearest shrunken centroids” identifies subsets of genes that best characterize each class, which was highly efficient in finding genes for classifying small round blue cell tumors and leukemias.
Journal of Bioinformatics and Computational BiologyMINIMUM REDUNDANCY FEATURE SELECTION FROM MICROARRAY GENE EXPRESSION DATA
2,794 Citations2005Chris Ding, HANCHUAN PENG
Journal of the American Statistical AssociationComparison of Discrimination Methods for the Classification of Tumors Using Gene Expression Data
2,699 Citations2002Sandrine Dudoit, Jane Fridlyand +1 more
Different discrimination methods for the classification of tumors based on gene expression data include nearest-neighbor classifiers, linear discriminant analysis, and classification trees, which are applied to datasets from three recently published cancer gene expression studies.
The Annals of StatisticsThe positive false discovery rate: a Bayesian interpretation and the q-value
2,305 Citations2003John D. Storey
This work introduces a modified version of the FDR called the “positive false discovery rate” (pFDR), which can be written as a Bayesian posterior probability and can be connected to classification theory.
Journal of the American Statistical AssociationEmpirical Bayes Analysis of a Microarray Experiment
1,758 Citations2001Bradley Efron, Robert Tibshirani +2 more
A simple nonparametric empirical Bayes model is introduced, which is used to guide the efficient reduction of the data to a single summary statistic per gene, and also to make simultaneous inferences concerning which genes were affected by the radiation.
Comprehensive Survey on Distance/Similarity Measures between Probability Density Functions
1,695 Citations2007Sung-Hyuk Cha
Various distance/similarity measures that are applicable to compare two probability density functions, pdf in short, are reviewed and categorized in both syntactic and semantic relationships to reveal similarities among numerous distance/Similarity measures.
BioinformaticsA Bayesian framework for the analysis of microarray expression data: regularized <i>t</i> -test and statistical inferences of gene changes
1,638 Citations2001Pierre Baldi, Anthony D. Long
A Bayesian probabilistic framework for microarray data analysis is developed that derives point estimates for both parameters and hyperparameters, and regularized expressions for the variance of each gene by combining the empirical variance with a local background variance associated with neighboring genes.
FEBS LettersRank products: a simple, yet powerful, new method to detect differentially regulated genes in replicated microarray experiments
1,521 Citations2004Rainer Breitling, Patrick Armengaud +2 more
An analysis of the physiological function of the identified genes indicates that the RP approach is powerful for identifying biologically relevant expression changes and can lead to a sharp reduction in the number of replicate experiments needed to obtain reproducible results.
Malay LiteratureWhat's In A Name? Malay Seals As Onomastic Sources
1,510 Citations2018Annabel Teh Gallop
STATISTICAL METHODS FOR IDENTIFYING DIFFERENTIALLY EXPRESSED GENES IN REPLICATED cDNA MICROARRAY EXPERIMENTS
1,414 Citations2002Sandrine Dudoit, Jean Yang +2 more
Differentially expressed genes are identified based on adjusted p-values for a multiple testing procedure which strongly controls the family-wise Type I error rate and takes into account the dependence structure between the gene expression levels.
Journal of Computational BiologyAnalysis of Variance for Gene Expression Microarray Data
1,325 Citations2000Kathleen F. Kerr, Mitchell Martin +1 more
It is demonstrated that ANOVA methods can be used to normalize microarray data and provide estimates of changes in gene expression that are corrected for potential confounding effects and establishes a framework for the general analysis and interpretation of micro array data.
Journal of Computational BiologyClustering Gene Expression Patterns
1,229 Citations1999Amir Ben‐Dor, Ron Shamir +1 more
Statistical ScienceMultiple Hypothesis Testing in Microarray Experiments
1,084 Citations2003Sandrine Dudoit, Juliet Popper Shaffer +1 more
Different approaches to multiple hypothesis testing in the context of DNA microarray experiments are discussed and a comparison of the procedures on microarray and simulated data sets is compared.
The LancetPrediction of cancer outcome with microarrays: a multiple random validation strategy
988 Citations2005Stefan Michiels, Serge Koscielny +1 more
Data is reanalysed from the seven largest published studies that have attempted to predict prognosis of cancer patients on the basis of DNA microarray analysis to study the stability of the molecular signature and the proportion of misclassifications.
Nature MethodsMultiple-laboratory comparison of microarray platforms
883 Citations2005Rafael A. Irizarry, Daniel Warren +23 more
A consortium of ten laboratories from the Washington, DC–Baltimore, USA, area was formed to compare data obtained from three widely used platforms using identical RNA samples to demonstrate that there are relatively large differences in data obtained in labs using the same platform, but that the results from the best-performing labs agree rather well.
BioinformaticsThe mutual information: Detecting and evaluating dependencies betweenvariables
834 Citations2002Ralf Steuer, Jürgen Kurths +3 more
The findings show that the algorithms used so far may be quite substantially improved upon when dealing with small datasets, finite sample effects and other sources of potentially misleading results have to be taken into account.
BioinformaticsOutcome signature genes in breast cancer: is there a unique set?
832 Citations2004Liat Ein‐Dor, Itai Kela +3 more
Molecular Systems BiologyHow to infer gene networks from expression profiles
781 Citations2007Mukesh Bansal, Vincenzo Belcastro +2 more
It is shown that reverse‐engineering algorithms are indeed able to correctly infer regulatory interactions among genes, at least when one performs perturbation experiments complying with the algorithm requirements.
Journal of Computational BiologyOn Differential Variability of Expression Ratios: Improving Statistical Inference about Gene Expression Changes from Microarray Data
757 Citations2001Michael A. Newton, Christina Kendziorski +3 more
This work considers the problem of inferring fold changes in gene expression from cDNA microarray data and derives estimates of gene expression changes within a simple hierarchical model that accounts for measurement error and fluctuations in absolute gene expression levels.
Journal of Computational BiologyTissue Classification with Gene Expression Profiles
743 Citations2000Amir Ben‐Dor, Laurakay Bruhn +4 more
Replicated microarray data
671 Citations2001Ingrid Lönnstedt
This paper presents an empirical Bayes method for analysing replicated microarray data and presents the results of a simulation study estimating the ROC curve of B and three other statistics for determining differential expression: the average and two simple modifications of the usual t-statistic.
Lecture notes in computer scienceRobust Feature Selection Using Ensemble Feature Selection Techniques
651 Citations2008Yvan Saeys, Thomas Abeel +1 more
It is shown that ensemble feature selection techniques show great promise for high-dimensional domains with small sample sizes, and provide more robust feature subsets than a single feature selection technique.
Feature selection for high-dimensional genomic microarray data
629 Citations2001Eric P. Xing, Michael I. Jordan +1 more
This paper reports on the successful application of feature selection methods to a classification problem in molecular biology involving only 72 data points in a 7130 dimensional space and investigates regularization methods as an alternative to feature selection.
Journal of BacteriologyFunctional Genomics: Expression Analysis of <i>Escherichia coli</i> Growing on Minimal and Rich Media
578 Citations1999Tao Han, Christoph Bausch +3 more
This experiment demonstrates that genomic expression assays can be applied in a meaningful way to the study of whole-bacterial-cell physiology for the generation of hypotheses and as a guide for more detailed studies of particular genes of interest.
Current BioinformaticsA Review of Ensemble Methods in Bioinformatics
510 Citations2010Pengyi Yang, Jean Yang +2 more
This article provides a review of the most widely used ensemble learning methods and their application in various bioinformatics problems, including the main topics of gene expression, mass spectrometry-based proteomics, gene-gene interaction identification from genome-wide association studies, and prediction of regulatory elements from DNA and protein sequences.
Computational Biology and ChemistryGene selection from microarray data for cancer classification—a machine learning approach
412 Citations2005Yu Wang, Igor V. Tetko +5 more
It is shown that classification performance at least as good as published results can be obtained on acute leukemia and diffuse large B-cell lymphoma microarray data sets and it is demonstrated that a combined use of different classification and feature selection approaches makes it possible to select relevant genes with high confidence.
PLoS ONEThe Influence of Feature Selection Methods on Accuracy, Stability and Interpretability of Molecular Signatures
382 Citations2011Anne-Claire Haury, Pierre Gestraud +1 more
It is observed that the feature selection method has a significant influence on the accuracy, stability and interpretability of signatures, and a simple Student's t-test seems to provide the best results.
Statistical Applications in Genetics and Molecular BiologyClassifying Gene Expression Profiles from Pairwise mRNA Comparisons
363 Citations2004Donald Geman, Christian d'Avignon +2 more
BMC BioinformaticsComparison and evaluation of methods for generating differentially expressed gene lists from microarray data
358 Citations2006Ian B. Jeffery, Desmond G. Higgins +1 more
Journal of Biological ChemistryImproved Statistical Inference from DNA Microarray Data Using Analysis of Variance and A Bayesian Statistical Framework
358 Citations2001Anthony D. Long, Harry J. Mangalam +4 more
New statistical tests based on thet test that can be conveniently used on high density array data to test for statistically significant differences between treatments identify genes that are up- or down-regulated following an experimental manipulation more reliably than approaches based only on a t test or fold change.
BioinformaticsClass prediction and discovery using gene microarray andproteomics mass spectroscopy data: curses, caveats, cautions
332 Citations2003Ray Somorjai, B. Dolenko +1 more
This work shows for several publicly available microarray and proteomics datasets how the 'curse of dimensionality' and dataset sparsity influence classification outcomes, and suggests an approach to assess the relative quality of apparently equally good classifiers.
Genome ResearchAn Efficient and Robust Statistical Modeling Approach to Discover Differentially Expressed Genes Using Genomic Expression Profiles
318 Citations2001Jeffrey G. Thomas, James M. Olson +2 more
A group of genes whose expression profiles correlated with that of thrombopoietin are identified and found that geneswhose expression associated with AML treatment outcome lie in recurrent chromosomal locations.
Genome biologyNew feature subset selection procedures for classification of expression profiles
263 Citations2002Trond Hellem Bø, Inge Jonassen
New methods for finding gene sets that are well suited for distinguishing experiment classes, such as healthy versus diseased tissues, are presented, based on evaluating genes in pairs and evaluating how well a pair in combination distinguishes two experiment classes.
Briefings in BioinformaticsStability and aggregation of ranked gene lists
191 Citations2009Anne‐Laure Boulesteix, Martin Slawski
This article reviews existing approaches for the assessment of stability of ranked gene lists and the related problem of aggregation, gives some practical recommendations, and warns against potential misuse of these methods.
BMC BioinformaticsAn Entropy-based gene selection method for cancer classification using microarray data
183 Citations2005Xiaoxing Liu, Arun V. Krishnan +1 more
An entropy-based iterative algorithm for selecting genes from microarray data that are able to classify various cancer sub-types with high accuracy is presented and the feature set obtained is very compact, that is, the redundancy between genes is reduced to a large extent, implying that classifiers can be built with a smaller subset of genes.
Functional & Integrative GenomicsA mixture model approach to detecting differentially expressed genes with microarray data
173 Citations2003Chap T. Le, Wei Pan +1 more
This work proposes a nonparametric statistical approach, called the mixture model method (MMM), to handle the problem when there are a small number of replicates under each experimental condition, and estimates the distributions of a t -type test statistic and its null statistic using finite normal mixture models.
Interface FocusHow to infer gene networks from expression profiles, revisited
173 Citations2011Christopher A. Penfold, David L. Wild
For smaller systems, DBNs are competitive with the non-parametric approaches with respect to computational time and accuracy, and both of these approaches appear to be more accurate than Granger causality-based methods and those using simple ODEs models.
Clustering gene expression patterns
165 Citations1999Amir Ben‐Dor, Zohar Yakhini
This paper defines an appropriate stochastic error model on the input, and proves that under the conditions of the model, the algorithm recovers the cluster structure with high probability, and presents a practical heuristic based on the same algorithmic ideas.
On the Consistency of Feature Selection using Greedy Least Squares Regression
162 Citations2009Tong Zhang
Genome biologyMulticlass classification of microarray data with repeated measurements: application to cancer
140 Citations2003Ka Yee Yeung, Roger E. Bumgarner
It is shown that removing highly correlated genes typically improves classification results using a small set of genes, and the uncorrelated shrunken centroid (USC) andEWUSC algorithms are developed.
A NONPARAMETRIC SCORING ALGORITHM FOR IDENTIFYING INFORMATIVE GENES FROM MICROARRAY DATA
121 Citations2000Peter J. Park, Marcello Pagano +1 more
A nonparametric scoring algorithm is introduced that assigns a score to each gene based on samples with known classes, and this procedure is robust to outliers and different normalization schemes, and immediately reduces the size of the data with little loss of information.
BioinformaticsEvaluating reproducibility of differential expression discoveries in microarray studies by considering correlated molecular changes
121 Citations2009Min Zhang, Lin Zhang +8 more
It was shown that though the POG scores for DEG lists from different studies for each disease are extremely low, the POGR and nPOGR scores can be rather high, suggesting that the apparently inconsistentDEG lists may be highly reproducible in the sense that they are actually significantly correlated.
BMC BioinformaticsGene selection for classification of microarray data based on the Bayes error
93 Citations2007Jigang Zhang, Hong‐Wen Deng
This study indicates that application of the Bayes error is a feasible and effective way for removing redundant genes in gene selection, and can effectively identify a compact set of genes with high classification accuracy.
BioinformaticsOn the use of permutation in and the performanceof a class of nonparametric methods to detectdifferential gene expression
92 Citations2003Wei Pan
A new method of constructing a null statistic that aims to estimate the null distribution of a test statistic directly directly is proposed and compared to the existing method and the new method when applied in EB, SAM and MMM.
Algorithms for Molecular BiologyRanking differentially expressed genes from Affymetrix gene expression data: methods with reproducibility, sensitivity, and specificity
85 Citations2009Koji Kadota, Yuji Nakai +1 more
This analysis of the MicroArray Quality Control (MAQC) project's datasets showed that the FC-based gene ranking methods (WAD, AD, FC, and RP) had a higher level of reproducibility: the percentages of overlapping genes (POGs) across different sites for the FC-based methods were higher overall than those for the t-statistic-based methods.
A rank sum test method for informative gene discovery
68 Citations2004Lin Deng, Jian Pei +2 more
This paper proposes a rank sum test method for informative gene discovery that uses a rank-sum statistic as the ranking criterion and proposes using the significance level threshold, instead of the number of informative genes, as the parameter.
Short term performance forecasting in enterprise systems
65 Citations2005Rob Powers, Moisés Goldszmidt +1 more
This study analyzes three dimensions that are important for their application as a stand alone tool and establishes that models induced using combined data from various systems generalize well and are applicable to new systems, enabling accurate predictions on systems with insufficient historical data.
BioinformaticsinSilicoDb: an R/Bioconductor package for accessing human Affymetrix expert-curated datasets from GEO
63 Citations2011Jonatan Taminau, David Steenhoff +9 more
The inSilicoDb R/Bioconductor package is a command-line front-end to the InSilico DB, a web-based database currently containing 86 104 expert-curated human Affymetrix expression profiles compiled from 1937 GEO repository series.
BioinformaticsNoise sampling method: an ANOVA approachallowing robust selection of differentially regulatedgenes measured by DNA microarrays
59 Citations2003Sorin Drăghici, Olga I. Kulaeva +4 more
The novel method proposed here is based on ANOVA and uses replicate spots to estimate an empirical distribution of the noise and suggests that the noise sampling method is likely to provide fewer false positives.
Journal of Bioinformatics and Computational BiologySIMILARITIES OF ORDERED GENE LISTS
58 Citations2006Xinan Yang, Stefan Bentink +2 more
A novel algorithm is presented to uncover that two lists of differentially expressed genes are similar, even if these similarities are not apparent to the eye, based on the ordering in the lists.
OMICS A Journal of Integrative BiologyA Two-Stage Feature Selection Method for Gene Expression Data
39 Citations2009Li‐Yeh Chuang, Chao-Hsuan Ke +2 more
This paper proposes a two-stage feature selection method, which uses information gain to implement a gene-ranking process, and combines an improved particle swarm optimization with the K-nearest neighbor method and support vector machine classifiers to calculate the classification accuracy.
Journal of Theoretical BiologyDetecting differentially expressed genes by relative entropy
34 Citations2005Xiting Yan, Minghua Deng +2 more
A new non-parametric method for selecting differentially expressed genes by relative entropy, called SDEGRE, to detect differentially expression genes by combining relative entropy and kernel density estimation, which can detect all types of differences between two groups of samples.
BMC BioinformaticsNot proper ROC curves as new tool for the analysis of differentially expressed genes in microarray experiments
26 Citations2008Stefano Parodi, Vito Pistoia +1 more
A new method for identifying differentially expressed genes, based on the area between the ROC curve and the rising diagonal (ABCR), which may correspond to those genes that tend to escape standard selection methods.
Lecture notes in computer scienceGene Selection for Cancer Classification through Ensemble of Methods
5 Citations2009Artur Wiliński, S. Osowski +1 more
The paper develops the methods of selection of the most important gene sequence on the basis of the gene expression microarray, corresponding to different types of cancer, and proposes special two stage strategy of selection.