Spectral biclustering of microarray data: coclustering genes and conditions.

By Yuval Kluger, Ronen Basri, Joseph T. Chang, Mark Gerstein. Genome Research, 2003

789 citations recorded by the source corpus.

Global analyses of RNA expression levels are useful for classifying genes and overall phenotypes. Often these classification problems are linked, and one wants to find "marker genes" that are differentially expressed in particular sets of "conditions." We have developed a method that simultaneously clusters genes and conditions, finding distinctive "checkerboard" patterns in matrices of gene expression data, if they exist. In a cancer context, these checkerboards correspond to genes that are markedly up- or downregulated in patients with particular types of tumors. Our method, spectral biclustering, is based on the observation that checkerboard structures in matrices of expression data can be found in eigenvectors corresponding to characteristic expression patterns across genes or conditions. In addition, these eigenvectors can be readily identified by commonly used linear algebra approaches, in particular the singular value decomposition (SVD), coupled with closely integrated normalization steps. We present a number of variants of the approach, depending on whether the normalization over genes and conditions is done independently or in a coupled fashion. We then apply spectral biclustering to a selection of publicly available cancer expression data sets, and examine the degree to which the approach is able to identify checkerboard structures. Furthermore, we compare the performance of our biclustering methods against a number of reasonable benchmarks (e.g., direct application of SVD or normalized cuts to raw data).

Subfields: Molecular Biology

View Paper: https://www.semanticscholar.org/paper/9cd32a77bfeaf797d378ce3c460acf8540916e42

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

Citations in the corpus

A causal framework for explaining the predictions of black-box sequence-to-sequence models (2017, methodology): “For the partitioning step we compare the robust formulation described above with two classical approaches to bipartite graph partitioning which do not take uncertainty into account: the coclustering method of Dhillon (2001) and the biclustering method of Kluger et al. (2003).”
FABIA: factor analysis for bicluster acquisition (2010, methodology): “A survey of biclustering approaches has been given by Madeira and Oliveira (2004). In principle, there exist four categories of biclustering methods: (1) variance minimization methods, (2) two-way clustering methods, (3) motif and pattern recognition methods and (4) probabilistic and generative appr”
The Inferelator: an algorithm for learning parsimonious regulatory networks from systems-biology data sets de novo (2006, background): “Biclustering was developed to address better the full complexity of finding co-regulated genes under multifactor control by grouping genes on the basis of coherence under subsets of observed conditions [10,16-22].”
Finding Correlated Biclusters from Gene Expression Data (2011, methodology): “Similar to spectral biclustering [ 46 ], [47], our method uses SVD to decompose the gene expression matrix into a group of basis genes and a group of basis conditions.”
A generalized maximum entropy approach to bregman co-clustering and matrix approximation (2004, background): “Currently, there exist many formulations of the co-clustering problem such as the hierarchical coclustering model (Hartigan, 1972), the sequential bi-clustering model (Cheng and Church, 2000) that involves finding the best co-clusters one at a time, and the spectral co-clustering model (Dhillon, 200”
QUBIC: a qualitative biclustering algorithm for analyses of gene expression data (2009, background): “Actually, there have been several studies that used biclustering as part of a larger analysis pipeline to do cancer subtyping (34).”
Finding large average submatrices in high dimensional data (2009, methodology): “G N ] 8 O ct 2 00 9 The Annals of Applied Statistics 2009, Vol. 3, No. 3, 985–1012 DOI: 10.1214/09-AOAS239 c© Institute of Mathematical Statistics, 2009”
Biclustering in data mining (2008, background): “Very similar spectral biclustering routines for microarray data have been suggested by Kluger et al. [25].”
Minimum Sum-Squared Residue Co-Clustering of Gene Expression Data (2004, methodology): “Different normalizations of genes and conditions are compared in [13]; however, the algorithms in [13] and [6, 20] model the gene expression matrix as a bipartite graph with non-negative edge weights where the quality of a co-cluster by the normalized cut criterion of [15].”
Data-Driven Tree Transforms and Metrics (2017, background): “The problem of matrix organization considered here is closely related to biclustering [1]–[5], [11], [12], where the goal is to identify biclusters: joint subsets of features and observations such that the matrix elements in each subset have similar values.”
Detecting unusual activity in video (2004, background): “If we think of the video segments as documents, and prototype features as keywords we observe that similar problems arise in the context of document-keyword clustering [8, 4].”
Biclustering algorithms for biological data analysis: a survey (2004, neutral): “The approaches followed by Busygin et al. [4], Kluget et al. [16] and Califano et al. [5] also discover all the biclusters at the same time.”
A survey of kernel and spectral methods for clustering (2008, methodology): “Spectral methods have been applied in clustering of artificial data [35] [36], in image segmentation [23] [37] [38], in bioinformatics [39], and in co-clustering problems of words and documents [40] and genes and conditions [41].”
Sparse multitask regression for identifying common mechanism of response to therapeutic targets (2010, methodology): “Co-clustering analysis has been used to find clusters in various tabulated data such as the co-occurrence of documents/words [4], or the expression of genes under various conditions [17, 5, 27], by simultaneously grouping rows and columns of the association table.”
PMLB: a large benchmark suite for machine learning evaluation and comparison (2017, methodology): “Model-dataset biclustering Figure 4 summarizes the results of biclustering the balanced accuracy of the tuned models according to the ML method and dataset using a spectral biclustering algorithm [27].”
PAC-Bayesian Analysis of Co-clustering and Beyond (2010, methodology): “This form of co-clustering is also known as partitional co-clustering (Banerjee et al., 2007), checkerboard bi-clustering (Cheng and Church, 2000; Kluger et al., 2003), grid clustering (Devroye et al., 1996; Seldin and Tishby, 2008, 2009), and box clustering.”