摘要：The weighted gene co-expression network analysis (WGCNA) has been used to explore gene expression datasets by constructing biological networks based on the likelihood expression profile among genes. In recent years, WGCNA found application in biomarker discovery studies, including miRNA. Serum samples from 20 patients with hepatocellular carcinoma (HCC) were profiled through miRNA 3.0 gene array and miRNAs biomarker candidates were identified through WGCNA. Results were validated by qRT-PCR in 102 HCC serum samples collected at diagnosis. WGCNA identified 16 miRNA modules, nine of them were significantly associated with the clinical characteristics of the patient. The Red module had a significant negative correlation with patients Survival (− 0.59, p = 0.007) and albumin (− 0.52, p = 0.02), and a positive correlation with PCR (0.61, p = 0.004) and alpha-fetoprotein (0.51, p = 0.02). In the red module, 16 circulating miRNAs were significantly associated with patient survival. MiR-3185 and miR-4507 were identified as predictors of patient survival after the validation phase. At diagnosis, high expression of circulating miR-3185 and miR-4507 identifies patients with longer survival (HR 2.02, 95% CI 1.10–3.73, p = 0.0086, and HR of 1.75, 95% CI 1.02–3.02, p = 0.037, respectively). Thought a WGCNA we identified miR-3185 and miR-4507 as promising candidate biomarkers predicting a longer survival in HCC patients.
其他摘要：Abstract The weighted gene co-expression network analysis (WGCNA) has been used to explore gene expression datasets by constructing biological networks based on the likelihood expression profile among genes. In recent years, WGCNA found application in biomarker discovery studies, including miRNA. Serum samples from 20 patients with hepatocellular carcinoma (HCC) were profiled through miRNA 3.0 gene array and miRNAs biomarker candidates were identified through WGCNA. Results were validated by qRT-PCR in 102 HCC serum samples collected at diagnosis. WGCNA identified 16 miRNA modules, nine of them were significantly associated with the clinical characteristics of the patient. The Red module had a significant negative correlation with patients Survival (− 0.59, p = 0.007) and albumin (− 0.52, p = 0.02), and a positive correlation with PCR (0.61, p = 0.004) and alpha-fetoprotein (0.51, p = 0.02). In the red module, 16 circulating miRNAs were significantly associated with patient survival. MiR-3185 and miR-4507 were identified as predictors of patient survival after the validation phase. At diagnosis, high expression of circulating miR-3185 and miR-4507 identifies patients with longer survival (HR 2.02, 95% CI 1.10–3.73, p = 0.0086, and HR of 1.75, 95% CI 1.02–3.02, p = 0.037, respectively). Thought a WGCNA we identified miR-3185 and miR-4507 as promising candidate biomarkers predicting a longer survival in HCC patients.
