Significant genes extraction and analysis of gene expression data based on matrix factorization techniques

Wei Kong, Juan Wang, Xiaoyang Mou

Research output: Contribution to journalArticlepeer-review

Abstract

It is generally considered that various regulatory activities between genes are contained in the gene expression datasets. Therefore, the underlying gene regulatory relationship and the biologically useful information can be found by modeling the gene regulatory network from the gene expression data. In our study, two unsupervised matrix factorization methods, independent component analysis (ICA) and nonnegative matrix factorization (NMF), were proposed to identify significant genes and model the regulatory network using the microarray gene expression data of Alzheimer's disease (AD). By bio-molecular analyzing of the pathways, the differences between ICA and NMF have been explored and the fact, which the inflammatory reaction is one of the main pathological mechanisms of AD, is also emphasized. It was demonstrated that our study gave a novel and valuable method for the research of early detection and pathological mechanism, biomarkers' findings of AD.

Original languageEnglish (US)
Pages (from-to)662-670
Number of pages9
JournalSheng wu yi xue gong cheng xue za zhi = Journal of biomedical engineering = Shengwu yixue gongchengxue zazhi
Volume31
Issue number3
StatePublished - Jun 1 2014
Externally publishedYes

All Science Journal Classification (ASJC) codes

  • General Medicine

Fingerprint

Dive into the research topics of 'Significant genes extraction and analysis of gene expression data based on matrix factorization techniques'. Together they form a unique fingerprint.

Cite this