GCBRGCN: Integration of ceRNA and RGCN to Identify Gastric Cancer Biomarkers.

Abstract

Gastric cancer (GC) is a prevalent malignancy, and the discovery of biomarkers plays a crucial role in the diagnosis and prognosis of GC. However, current strategies for identifying GC biomarkers often focus on a single ribonucleic acid (RNA) class, neglecting the potential for multiple RNA types to collectively serve as biomarkers with improved predictive capabilities. To bridge this gap, our study introduces the GC biomarker relation graph convolution neural network (GCBRGCN) model which integrates the competing endogenous RNA (ceRNA) network with GC clinical informations and whole transcriptomics data, leveraging the relational graph convolutional network (RGCN) to predict GC biomarkers. It demonstrates exceptional performance, surpassing traditional machine learning and graph neural network algorithms with an area under the curve (AUC) of 0.8172 in the task of predicting GC biomarkers. Our study identified three unreported potential novel GC biomarkers: CCNG1, CYP1B1, and CITED2. Moreover, FOXC1 and LINC00324 were characterized as biomarkers with significance in both prognosis and diagnosis. Our work offers a novel framework for GC biomarker identification, highlighting the critical role of multiple types RNA interaction in oncological research.