Integrated Multiple “-omics” Data Reveal Subtypes of Hepatocellular Carcinoma

• DOI: 10.1371/journal.pone.0165457
• Publication Date: 2016-11-02T17:52:11Z
• Authors: Gang Liu; Chuanpeng Dong; Lei Liu
• Keywords: False Discovery Rate

Hepatocellular carcinoma is one of the most heterogeneous cancers, as reflected by its multiple grades and difficulty to subtype. In this study, we integrated copy number variation, DNA methylation, mRNA, miRNA data with developed "cluster cluster" method classified 256 HCC samples from TCGA (The Cancer Genome Atlas) into five major subgroups (S1-S5). We observed that classification was associated specific mutations protein expression, detected each subgroup had distinct molecular signatures. The subclasses were not only survival but also clinical observations. S1 characterized bulk amplification on 8q24, TP53 mutation, low lipid metabolism, highly expressed onco-proteins, attenuated tumor suppressor proteins a worse rate. S2 S3 telomere hypomethylation expression TERT DNMT1/3B. Compared S2, less variation some good prognosis biomarkers, including CRP CYP2E1. contrast, mutation rate CTNNB1 higher in S3. S4 various characteristics at different biological levels. summary, using "-omics" data. Each signature, which may provide information about pathogenesis subtypes HCC.