An Integrated Graph Regularized Non-Negative Matrix Factorization Model for Gene Co-Expression Network Analysis

Studies of cancers have become diversified in recent years, especially with the availability of multi-omics data. Establishing an effective integrative model to process more types of data has become a new research hotspot. In order to conduct deeper mining in cancer, building a gene co-expression ne...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE access 2019, Vol.7, p.126594-126602
Hauptverfasser: Gao, Ying-Lian, Hou, Mi-Xiao, Liu, Jin-Xing, Kong, Xiang-Zhen
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Studies of cancers have become diversified in recent years, especially with the availability of multi-omics data. Establishing an effective integrative model to process more types of data has become a new research hotspot. In order to conduct deeper mining in cancer, building a gene co-expression network based on multi-omics data for more valuable clues is the useful means in this research. Based on the data-driven problems of cancer networks, this paper proposes an integrated Graph Regularized Non-negative Matrix Factorization model that can be used for network analysis called iGMFNA. We apply iGMFNA to two cancer datasets from The Cancer Genome Atlas (TCGA) for analysis. We demonstrate that our method is indeed more effective than other integrated methods. In terms of network analysis and mining, we also define a multi-measure for nodes in the network to identify cancer-related genes. Through text mining, we verify some genes discovered by iGMFNA.
ISSN:2169-3536
2169-3536
DOI:10.1109/ACCESS.2019.2939405