Coupled two-way clustering analysis of data.

• US 20050240563A1
• Filed: 06/17/2005
• Published: 10/27/2005
• Est. Priority Date: 03/09/2000
• Status: Active Grant

• Current Assignee: Yeda Research and Development Co. Ltd. (Weizmann Institute of Science)
• Original Assignee: Yeda Research and Development Co. Ltd. (Weizmann Institute of Science)
• Inventors: Levine, Erel, Getz, Gad, Domany, Eytan
• Granted Patent: US 7,599,933 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 18/231 Hierarchical techniques, i.e. dividing or merging pattern sets so as to obtain a dendrogram
• CPC Class Codes: G16B 25/00 ICT specially adapted for hybridisation; ICT specially adapted for gene or protein expression
• CPC Class Codes: G16B 25/10 Gene or protein expression profiling; Expression-ratio estimation or normalisation
• CPC Class Codes: G16B 25/30 Microarray design
• CPC Class Codes: G16B 40/00 ICT specially adapted for biostatistics; ICT specially adapted for bioinformatics-related machine learning or data mining, e.g. knowledge discovery or pattern finding
• CPC Class Codes: G16B 40/30 Unsupervised data analysis
• CPC Class Codes: Y10S 707/99937 Sorting

Abstract

A novel coupled two-way clustering approach to gene microarray data analysis, for identifying subsets of the genes and samples, such that when one of these items is used to cluster the other, stable and significant partitions emerge. The method of the present invention preferably uses iterative clustering in order to execute this search in an efficient way. This approach is especially suitable for gene microarray data, where the contributions of a variety of biological mechanisms to the gene expression levels are entangled in a large body of experimental data. The method of the present invention was applied to two gene microarray data sets, on colon cancer and leukemia. By identifying relevant subsets of the data and focusing on these subsets, partitions and correlations were found that were masked and hidden when the full data set was used in the analysis.

Claims

1-25. -25. (canceled)

26. A method for analyzing data, available in the form of an array of numbers, wherein each row of the array represents measurements of the values taken by a particular attribute over several samples and each column represents the measurements of the various attributes taken for a particular sample.

27. The method according to claim 26, for the analysis of gene expression data taken from several tissues, wherein the attributes are different genes for which expression levels were measured and the samples are human tissues or other biological material for which the expression levels of the genes have been determined.

28. The method according to claim 26, comprising performing cluster analysis in two ways, over the samples and over the genes, wherein the two ways of clustering are coupled:

each cluster of genes constitutes a probe which is used to cluster any group of samples, and vice versa.

29. The method according to claim 26, wherein the method is iterative and whenever stable clusters are generated, they are used to further search for partitions (clusters) in the other dimension.

30. The method according to claim 29, wherein gene clusters are used to look for partitions of tissues and tissue clusters are used to look for correlated clusters of genes.

31. The method according to claim 26, wherein the method is used in conjunction with any clustering algorithm.

32. The method according to claim 26, wherein the method is used in conjunction with the superparamagnetic clustering algorithm.

33. The method according to claim 32, which utilizes a measure for the stability of clusters and identification of said stable clusters narrows significantly the groups (clusters) that are to be tested as probes.

34. The method according to claim 26, which yields clusters of genes of correlated expression profiles that may participate in the same biological process.

35. The method according to claim 34, wherein said groups of genes relate to administration of pharmaceutical drugs, or differentiate one type of cancer from another, or reflect the change of experimental protocol in a colon-cancer treatment.

36. The method according to claim 35, wherein said method identifies tissues of groups of patients, or tissues subjected to different experimental protocols, or identifies different types of cancer.

37. The method according to claim 36, wherein said method identifies different types of leukemia.

38. (canceled)