Identification of Bicluster Regions in a Binary Matrix and Its Applications

Abstract

Biclustering has emerged as an important approach to the analysis of large-scale datasets. A biclustering technique identifies a subset of rows that exhibit similar patterns on a subset of columns in a data matrix. Many biclustering methods have been proposed, and most, if not all, algorithms are developed to detect regions of “coherence” patterns. These methods perform unsatisfactorily if the purpose is to identify biclusters of a constant level. This paper presents a two-step biclustering method to identify constant level biclusters for binary or quantitative data. This algorithm identifies the maximal dimensional submatrix such that the proportion of non-signals is less than a pre-specified tolerance δ. The proposed method has much higher sensitivity and slightly lower specificity than several prominent biclustering methods from the analysis of two synthetic datasets. It was further compared with the Bimax method for two real datasets. The proposed method was shown to perform the most robust in terms of sensitivity, number of biclusters and number of serotype-specific biclusters identified. However, dichotomization using different signal level thresholds usually leads to different sets of biclusters; this also occurs in the present analysis.

Application of biclustering algorithm in adverse drug reaction monitoring system of China

“…Bimax is designed to find the inclusion‐maximal submatrices (biclusters) of the data matrix in which all elements have a value equal to 1. Bimax is able to find the largest clusters in an ADR data, which is made of a subset of drugs that are all strongly associated with the same subset of AEs . Therefore, this analysis required the strength of the IC association to be dichotomized into “strong association” or “weak association,” respectively, with the “association strength threshold.” Additionally, Bimax requires the user to prespecify 2 parameters, namely, the minimum numbers of columns (drugs) and rows (AEs) needed to define a minimum bicluster.…”

A Composite Model for Subgroup Identification and Prediction via Bicluster Analysis

“…Consider a two-way data matrix with rows representing the measured attributes and columns representing samples. Many singular value decomposition (SVD) approaches for bicluster analysis of microarray data have been proposed and demonstrated to be effective [34] – [39] . In this paper, a SVD-based biclustering method [39] was used to identify substructures between subsets of attributes and subsets of samples.…”

“…Many singular value decomposition (SVD) approaches for bicluster analysis of microarray data have been proposed and demonstrated to be effective [34] – [39] . In this paper, a SVD-based biclustering method [39] was used to identify substructures between subsets of attributes and subsets of samples. An advantage of SVD-based biclustering methods is that the biclustering results do not depend on the random starting seeds.…”

“…Biclustering analysis provides an approach to identify substructures in the sampled data. Biclustering techniques identify biclusters by simultaneously clustering both samples and attributes [25] – [39] . Each bicluster is defined as a subset of attributes associated with a subset of samples.…”

“…Any biclustering methods can be used in the first step of bicluster identification. This paper uses a SVD-based biclustering algorithm to identify constant biclusters [39] ; this method has been shown to perform well in extensive comparisons with various biclustering methods, and found to be generally superior in terms of sensitivity and specificity. The primary focus of this paper is subgroup classification and prediction.…”