Metabolic Biomarkers For Cancers:A Bioinformatics Study

Metabolism is becoming a hot topic in recent cancer researches.A large number of metabolic biomarkers have been reported in published literatures.Hence building a comprehensive database for cancer metabolic biomarkers is propitious to the systems-level analysis of metabolic characteristics of cancers,which would promote the discovery of novel methods for cancer prevention and management.In this paper,we first constructed a database called CMBD(A Database of Metabolic Biomarkers for human Cancers,http://www.sysbio.org.cn/CMBD/)for cancer metabolic biomarkers.We downloaded published literatures from PubMed based on specific keywords,and extracted the information according to standardized annotation principles.Finally,a total of 358 "biomarker-cancer-reference" associations,including 225 biomarkers and 73 cancer types from 200 literatures,were manually recorded,and an interactive database with user-friendly web interfaces was developed.Based on this,we investigated the metabolic characteristics of breast cancer and prostate cancer through ftunctional enrichment analysis,biological network analysis,and gene expression analysis,respectively.The results showed that eight pathways including glycolysis etc.were closely related to breast cancer.Meanwhile,35 metabolic enzymes including ENO1 etc.were located in key sites in the network,22 metabolic genes including FBP1 etc.were significantly differentially expressed in breast cancer datasets,and five genes,i.e.,ACSS2,CAT,GPD1,GPD1L,MBOAT2,could be served as potential biomarkers for breast cancer.In the aspect of prostate cancer,11 metabolic pathways including citric acid cycle etc.were significantly enriched,43 metabolic enzymes were located in key sites in the network,34 genes including MCCC2 etc.were differentially expressed,two genes,i.e.,FH,ALDH2,could be considered as candidate biomarkers.This part demonstrated cancer heterogeneity at the level of metabolic pathways,protein interaction network and differentially expressed genes,respectively,and indicated the effects of metabolic changes on cancer development.This paper performed bioinformatics analysis on cancer metabolic biomarkers at the systems level,which is of theoretical significance and clinical value.