DFASGCNS: A prognostic model for ovarian cancer prediction based on dual fusion channels and stacked graph convolution

• DOI: 10.1371/journal.pone.0315924
• Publication Date: 2024-12-16T18:33:11Z
• Authors: Huiqing Wang; Xiao Han; Shuaijun Niu; Hao Cheng; Jianxue Ren; Yimeng Duan
• Keywords: Omics

Ovarian cancer is a malignant tumor with different clinicopathological and molecular characteristics. Due to its nonspecific early symptoms, the majority of patients are diagnosed local or extensive metastasis, severely affecting treatment prognosis. The occurrence ovarian influenced by multiple complex mechanisms including genomics, transcriptomics, proteomics. Integrating types omics data aids in predicting survival rate patients. However, existing methods only fuse multi-omics at feature level, neglecting shared complementary neighborhood information among samples data, failing consider potential interactions between level. In this paper, we propose prognostic model for prediction named Dual Fusion Channels Stacked Graph Convolutional Neural Network (DFASGCNS). DFASGCNS utilizes dual fusion channels learn representations associations samples. graph convolutional network used comprehensively deep intricate correlation networks present enhancing model’s ability represent data. An attention mechanism introduced allocate weights important features optimizing representation Experimental results demonstrate that compared methods, exhibits significant advantages prognosis analysis. Kaplan-Meier curve analysis indicate differences subgroups predicted model, contributing deeper understanding pathogenesis providing more reliable auxiliary diagnostic assessment