Abstract
Endometrial cancers (ECs) are the most common gynecologic malignancy. Many studies have investigated the tumorigenesis process and identified the biomarkers for signature classification. However, to the best of knowledge, there is no satisfactory gene network-based method on the investigation of tumorigenesis characterization and diagnosis of subtype of ECs. An integrative analysis of weighted gene co-expression network (WGCNA) and elastic-net regression model was used to reconstruct an ECs-specific gene co-expression network and to discover the cancer gene signatures driving tumorigenesis of ECs. This study reconstructed a co-expression network consisted of 620 genes, and discovered 19-gene diagnostic biomarker signatures for grade, type and stage in the EC-specific network contributing to the ECs progression via cell-cycle regulation, antigen processing and the citric acid (TCA) cycle. The predictive model achieved the good classifier performance for phenotypic characteristics of ECs. On the other hand, this study also applied the network-based analysis to a case study of bisphenol A (BPA) toxicity testing. BPA can play as an endocrine disrupting chemical and is usually used as sealants in plastic materials. The primary route of human exposure to BPA is food intake. This study presented the toxicity evaluation of BPA exposure in the visualized perturbations of molecular and cellular changes. This study provided a powerful biomarker discovery platform to better understand the progression of ECs to uncover potential therapeutic targets in the treatment of ECs, and the toxicity of environmental BPA exposure for health effect estimation and prevention.