Using multiobjective optimization for biclustering microarray data

[Display omitted] •A new multiobjective modeling for the biclustering problem.•A new hybrid multiobjective algorithm gradually conceived for best results.•Extracting relevant biclusters with large sizes compared to classical methods. Microarray data analysis is a challenging problem in the data mini...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Applied soft computing Ročník 33; s. 239 - 249
Hlavní autoři: Seridi, Khedidja, Jourdan, Laetitia, Talbi, El-Ghazali
Médium: Journal Article
Jazyk:angličtina
Vydáno: Elsevier B.V 01.08.2015
Elsevier
Témata:
Biclustering problem
Combinatorics
Evolutionary algorithm
Gene expression data
Mathematics
Multiobjective combinatorial optimization
Biclustering problem
Multiobjective combinatorial optimization
Gene expression data
Evolutionary algorithm
ISSN:1568-4946, 1872-9681
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:[Display omitted] •A new multiobjective modeling for the biclustering problem.•A new hybrid multiobjective algorithm gradually conceived for best results.•Extracting relevant biclusters with large sizes compared to classical methods. Microarray data analysis is a challenging problem in the data mining field. Actually, it represents the expression levels of thousands of genes under several conditions. The analysis of this data consists on discovering genes that share similar expression patterns across a sub-set of conditions. In fact, the extracted informations are submatrices of the microarray data that satisfy a coherence constraint. These submatrices are called biclusters, while the process of extracting them is called biclustering. Since its first application to the analysis of microarray [1], many modeling and algorithms have been proposed to solve it. In this work, we propose a new multiobjective model and a new metaheuristic HMOBIibea for the biclustering problem. Results of the proposed method are compared to those of other existing algorithms and the biological relevance of the extracted information is validated. The experimental results show that our method extracts very relevant biclusters, with large sizes with respect to existing methods.
ISSN:1568-4946
1872-9681
DOI:10.1016/j.asoc.2015.03.060