Development of a 12-biomarkers-based prognostic model for pancreatic cancer using multi-omics integrated analysis

Yanhui  Jia; Meiyan  Shen; Yan  Zhou; Huaiping Liu

doi:10.18388/abp.2020_5225

Yanhui Jia Laboratory of Clinical Biochemistry of Tianjin Third Central Hospital, Key Laboratory of Artificial Cell, Institute of Hepatobiliary Disease, Tianjin City, 300170, China https://orcid.org/0000-0002-3971-0769
Meiyan Shen Laboratory of Clinical Biochemistry, Tianjin Medical University, Tianjin City, 300070, China https://orcid.org/0000-0003-4223-3243
Yan Zhou Laboratory of Clinical Biochemistry, Tianjin Medical University, Tianjin City, 300070, China https://orcid.org/0000-0002-0127-0956
Huaiping Liu Laboratory of Clinical Biochemistry of Tianjin Third Central Hospital, Key Laboratory of Artificial Cell, Institute of Hepatobiliary Disease, Tianjin City, 300170, China https://orcid.org/0000-0003-3955-373X

DOI: https://doi.org/10.18388/abp.2020_5225

Abstract

Pancreatic cancer is one of the most malignant tumors of the digestive system, with insidious, rapid onset and high mortality. The 5-year survival rate is only 10%. Therefore, in-depth exploration of the potential mechanism affecting the prognosis of pancreatic cancer, and search for biomarkers that can effectively predict the prognosis of pancreatic cancer are of practical clinical importance. The mRNA sequencing data, miRNA sequencing data, methylation data and SNP data of pancreatic cancer patients available in The Cancer Genome Atlas (TCGA) were used for analysis to identify biomarkers that significantly affect the prognosis for the patients. Finally, a prognostic prediction model was developed using principal component analysis (PCA) method. The genes that significantly affected the prognosis of pancreatic cancer were as follows: 5 DmiRNAs (hsa-mir-1179, hsa-mir-1224, hsa-mir-1251, hsa-mir-129-1 and hsa-mir-129-2), 6 DmRNAsandDMsandMethyCor database entries (MAPK8IP2, CPE, DPP6, MSI1, IL20RB and S100A2), and FMN2 gene from differential expressed mRNAs and differential single-nucleotide polymorphism (DmRNAsandDSNPs) database. Prognostic index (PI)=∑iwi xi – 0.717716. A patient was predicted as high/low risk if the PI was larger/smaller than 0.034045. Our study resulted in a comprehensive prognostic model for pancreatic cancer patients based on multi-omics analysis, which could offer better guidance for the clinical management of patients with early-stage pancreatic cancer.