Triclustering analysis using extended dimension iterative signature algorithm (EDISA) on lung disease gene expression data

Dwi Aji Apriana, Titin Siswantining, Devvi Sarwinda, Saskya Mary Soemartojo

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Triclustering has been applied to three-dimensional gene expression data (gene, condition, and time) to group the dataset into sub-matrix groups that have similarities. One of the algorithms of triclustering analysis is the Extended Dimension Iterative Signature Algorithm (EDISA). This algorithm considers the Pearson distance between each gene and condition against the mean vector as a measure of similarity. The primary process in EDISA is the iteration process by deleting each gene and condition with a Pearson distance to the mean vector above a certain threshold. It is a measure of the similarity of a gene and condition to the mean of the tricluster candidate. EDISA was applied for lung disease gene expression's data using several scenarios with different thresholds. The result is that the higher the threshold value of each gene and condition, the more genes and conditions in the tricluster. Also, an evaluation was carried out using the Tricluster Diffusion (TD) Score value to find the best scenario where the best scenario was the scenario with the smallest TD Score. This algorithm's application to lung disease data generates triclusters, which can detect genes that distinguish the characteristics of patients with lung disease and healthy patients.

Original languageEnglish
Title of host publicationIBIOMED 2020 - Proceedings of the 37th International Conference on Biomedical Engineering
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages7-12
Number of pages6
ISBN (Electronic)9781728171562
DOIs
Publication statusPublished - 6 Oct 2020
Event37th International Conference on Biomedical Engineering, IBIOMED 2020 - Yogyakarta, Indonesia
Duration: 6 Oct 20208 Oct 2020

Publication series

NameIBIOMED 2020 - Proceedings of the 37th International Conference on Biomedical Engineering

Conference

Conference37th International Conference on Biomedical Engineering, IBIOMED 2020
Country/TerritoryIndonesia
CityYogyakarta
Period6/10/208/10/20

Keywords

  • EDISA
  • Gene Expression Data
  • Pearson Distance
  • Threshold Value
  • Tricluster Diffusion Score

Fingerprint

Dive into the research topics of 'Triclustering analysis using extended dimension iterative signature algorithm (EDISA) on lung disease gene expression data'. Together they form a unique fingerprint.

Cite this