Gene co-expression network analysis of transcriptomics data of gastric cancer subtypes to explore subtype-specific significant biological processes

Gastric cancer (GC) is one of the most prevalent cancers worldwide that has heterogeneous nature from pathological and molecular standpoints. So this is important to identify involved mechanisms in incidence and progression of GC. High throughput techniques allow studying molecular profiles of this disease. Understanding the gene expression regulation mechanism can be important to declare complex phenotypes and mainly the causes of cancers’ occurrence. The key step in propelling GC therapy toward personalized medicine is a classification of tumors in different subtypes based on different characteristics. One of these classifications has been done by Asian Cancer Research Group who used gene expression data to make 4 molecular subtypes (MSI, MSS/EMT, MSS/Tp53+, MSS/Tp53-). Due to the huge number of variables in transcriptomics data analysis, one of the best solutions is using network-based models. Gene networks can be used to studying gene regulatory mechanisms. In gene co-expression networks, expression correlation of genes is used to construct a gene network. We have applied the weighted gene co-expression network to create co-expressed gene modules for GC subtypes. In the next step, functional annotation analysis using DAVID web-server was used to identify involved biological pathways for the explored modules. The important modules that involved in biological pathways including 8, 10, 10, 9 co-expressed modules were identified and presented for MSI, EMT, Tp53+, Tp53- subtypes respectively. The extracted important pathways for each module are presenting as new and useful strategies for targeted treatment of GC.

Keywords:

Gastric cancer , Transcriptomics data , Weighted gene co-expression network analysis , Personalized medicine