Uncovering the potential differentially expressed miRNAs as diagnostic biomarkers for hepatocellular carcinoma based on machine learning in The Cancer Genome Atlas database

Background/Aims: Few studies have been designed to directly investigate the exact mechanisms underlying the different phenotypes between cirrhotic and non-cirrhotic hepatocellular carcinoma (HCC). This study aimed to illuminate the incidence and developmental mechanisms for both types of HCC through differentially expressed microRNAs (miRNAs) using bioinformatic analysis. Methods: The miRNA-seq data and clinical data of patients (from The Cancer Genome Atlas (TCGA) database) were utilized to determine differentially expressed miRNAs between cirrhotic and non-cirrhotic HCC. Afterwards, the function of the selected miRNAs was predicted with online tools. Furthermore, the miRNA expression and clinical significance were validated by external datasets and our experiment. Results: The present study included 135 non-cirrhotic and 80 cirrhotic patients to find differentially expressed miRNAs. Among them, four miRNAs (mir-1296, mir-23c, mir-149, and mir-95) were finally selected for further validation and analysis. Correlation analysis found that two miRNAs are greatly associated with the non-cirrhotic HCC patients’ clinical traits, such as the T, M, and N tumor stages. However, only mir-23c was associated with the cirrhotic HCC patients’ tumor T and N stages. Furthermore, survival analysis revealed that increased mir-149 in non-cirrhotic HCC, patients’ age and the existence of vessels in the tumor in cirrhotic HCC were independent risk factors for the patients’ postoperative overall survival. Functional and interaction analyses also supported the notion that these miRNAs function through some pathways that influence the behavior of the HCC. These results are strongly supported by analysis of extra datasets and our experiment. Conclusions: The cirrhotic and non-cirrhotic HCC types involve differentially expressed miRNAs, which are correlated with distinctive pathological traits. To some extent, non-cirrhotic HCC seems more dependent on miRNA network regulation, but additional studies are needed to confirm this conclusion.

Download Full-text

A Primer on a Hepatocellular Carcinoma Bioresource Bank Using the Cancer Genome Atlas Guidelines: Practical Issues and Pitfalls

World Journal of Surgery ◽

10.1007/s00268-010-0953-y ◽

2011 ◽

Vol 35 (8) ◽

pp. 1732-1737 ◽

Cited By ~ 1

Author(s):

N. Thao T. Nguyen ◽

Ron T. Cotton ◽

Theresa R. Harring ◽

Jacfranz J. Guiteau ◽

Marie-Claude Gingras ◽

...

Keyword(s):

Hepatocellular Carcinoma ◽

Cancer Genome ◽

The Cancer Genome Atlas ◽

Cancer Genome Atlas ◽

Genome Atlas

Download Full-text

Peer Review #1 of "Systematic analysis and prediction model construction of alternative splicing events in hepatocellular carcinoma: a study on the basis of large-scale spliceseq data from The Cancer Genome Atlas (v0.1)"

10.7287/peerj.8245v0.1/reviews/1 ◽

2019 ◽

Keyword(s):

Hepatocellular Carcinoma ◽

Alternative Splicing ◽

Prediction Model ◽

Peer Review ◽

Large Scale ◽

Cancer Genome ◽

The Cancer Genome Atlas ◽

Systematic Analysis ◽

Cancer Genome Atlas ◽

Genome Atlas

Download Full-text

Identification and Potential Mechanisms of a 7-MicroRNA Signature That Predicts Prognosis in Patients with Lower-Grade Glioma

Journal of Healthcare Engineering ◽

10.1155/2021/3251891 ◽

2021 ◽

Vol 2021 ◽

pp. 1-11

Author(s):

Zhizheng Liu ◽

Hongliang Meng ◽

Miaoxian Fang ◽

Wenlong Guo

Keyword(s):

Signaling Pathway ◽

Predictive Accuracy ◽

Cancer Genome ◽

The Cancer Genome Atlas ◽

Differentially Expressed ◽

Lower Grade ◽

Clinical Practices ◽

Lower Grade Glioma ◽

Cancer Genome Atlas ◽

Genome Atlas

Background. Lower-grade glioma is an intracranial cancer that may develop into glioblastoma with high mortality. The main objective of our study is to develop microRNA for LGG patients which will provide novel prognostic biomarkers along with therapeutic targets. Methods. Clinicopathological data of LGG patients and their RNA expression profile were downloaded through The Cancer Genome Atlas Relevant expression profiles of RNA, and clinicopathological data of the LGG patients had been extracted from the database of “The Cancer Genome Atlas.” Differential expression analysis had been conducted for identification of the differentially expressed microRNAs as well as mRNAs in LGG samples and normal ones. ROC curves and K–M plots were plotted to confirm performance and for predictive accuracy. For the confirmation of microRNAs as an independent prognostic factor, an independent prognosis analysis was conducted. Moreover, target differentially expressed genes of these identified prognostic microRNAs that were extracted and protein-protein interaction networks were developed. Moreover, the biological functions of signature were determined through Genome Ontology analysis, genome pathway analysis, and Kyoto Encyclopedia of Genes. Results. 7-microRNA signature was identified that has the ability of categorization of individuals with LGG into high- and low-risk groups on the basis of significant difference in survival during training and testing cohorts (P < 0.001). The 7-microRNA signature had appeared to be robust in predictive accuracy (all AUC> 0.65). It was also approved with multivariate Cox regression along with some traditional clinical practices that we can use 7-microRNA signature for therapeutic purposes as a self-regulating predictive OS factor (P < 0.001). KEGG and Gene Ontology (GO) analyses reported that 7-microRNAs had mainly developed in important pathways related with glioma, e.g., the “cAMP signaling pathway,” “glutamatergic synapses,” and “calcium signaling pathway”. Conclusion. A newly discovered 7-microRNA signature could be a potential target for the diagnosis and treatment for LGG patients.

Download Full-text