Title: Finding correlated biclusters from microarray data using the modified lift algorithm based on new residue score

Authors: Alhadi Bustamam; Soeganda Formalidin; Titin Siswantining; Zuherman Rustam

Addresses: Department of Mathematics, FMIPA Universitas Indonesia, Depok 16424, Jakarta, Indonesia ' Department of Mathematics, FMIPA Universitas Indonesia, Depok 16424, Jakarta, Indonesia ' Department of Mathematics, FMIPA Universitas Indonesia, Depok 16424, Jakarta, Indonesia ' Department of Mathematics, FMIPA Universitas Indonesia, Depok 16424, Jakarta, Indonesia

Abstract: The purpose of this research is to find a strong correlation between genes and conditions of diabetes mellitus gene expression data from obese and lean people using three-phase biclustering. The first step is to use Singular Value Decomposition (SVD) to decompose matrix gene expression data into two global-based gene and condition matrices. The second step is to use Partition around Medoid (PAM) to cluster gene and condition-based matrices using Euclidean distance, forming several biclusters that were further evaluated using the Modified Lift Algorithm based on Pearson correlation, which is a very appropriate method to detect an additive-multiplicative bicluster type. The algorithm processes are run using open-source R software. The resulting biclusters of the proposed algorithm having a strong correlation among genes and samples are obtained so that the method has high potential in future medical research.

Keywords: correlated bicluster; diabetes mellitus; microarray data; MLA; modified lift algorithm; R software.

DOI: 10.1504/IJDMB.2020.113691

International Journal of Data Mining and Bioinformatics, 2020 Vol.24 No.4, pp.326 - 343

Accepted: 14 Sep 2020
Published online: 18 Mar 2021 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article