Identification of a Novel Prognostic Signature for Gastric Cancer Based on Multiple Level Integration and Global Network Optimization

Lin Cui; Ping Wang; Dandan Ning; Jing Shao; Guiyuan Tan; Dajian Li; Xiaoling Zhong; Wanqi Mi; Chunlong Zhang; Shizhu Jin

doi:10.3389/fcell.2021.631534

Identification of a Novel Prognostic Signature for Gastric Cancer Based on Multiple Level Integration and Global Network Optimization

Front Cell Dev Biol. 2021 Apr 12:9:631534. doi: 10.3389/fcell.2021.631534. eCollection 2021.

Authors

Lin Cui¹, Ping Wang², Dandan Ning¹, Jing Shao¹, Guiyuan Tan³, Dajian Li⁴, Xiaoling Zhong³, Wanqi Mi³, Chunlong Zhang³, Shizhu Jin¹

Affiliations

¹ Department of Gastroenterology and Hepatology, The Second Affiliated Hospital, Harbin Medical University, Harbin, China.
² Department of Interventional Radiology, The Third Affiliated Hospital, Harbin Medical University, Harbin, China.
³ College of Bioinformatics Science and Technology, Harbin Medical University, Harbin, China.
⁴ Department of Gastroenterology and Hepatology, The First Hospital Of Harbin, Harbin, China.

Abstract

Gastric Cancer (GC) is a common cancer worldwide with a high morbidity and mortality rate in Asia. Many prognostic signatures from genes and non-coding RNA (ncRNA) levels have been identified by high-throughput expression profiling for GC. To date, there have been no reports on integrated optimization analysis based on the GC global lncRNA-miRNA-mRNA network and the prognostic mechanism has not been studied. In the present work, a Gastric Cancer specific lncRNA-miRNA-mRNA regulatory network (GCsLMM) was constructed based on the ceRNA hypothesis by combining miRNA-target interactions and data on the expression of GC. To mine for novel prognostic signatures associated with GC, we performed topological analysis, a random walk with restart algorithm, in the GCsLMM from three levels, miRNA-, mRNA-, and lncRNA-levels. We further obtained candidate prognostic signatures by calculating the integrated score and analyzed the robustness of these signatures by combination strategy. The biological roles of key candidate signatures were also explored. Finally, we targeted the PHF10 gene and analyzed the expression patterns of PHF10 in independent datasets. The findings of this study will improve our understanding of the competing endogenous RNA (ceRNA) regulatory mechanisms and further facilitate the discovery of novel prognostic biomarkers for GC clinical guidelines.

Keywords: competing endogenous RNAs; gastric cancer; integrated analysis; prognostic signature; random walk algorithm.