Identification Of The Hub Genes On Notch And HIF Signaling Pathway Associated With The Diagnosis And Prognosis Of Hepatocellular Carcinoma

Objective:Hepatocellular carcinoma(HCC)is the sixth most commonly diagnosed cancer,with more than 800,000 newly diagnosed cases and over 400,000 related deaths per year.Although the treatment of HCC has developed rapidly in recent years,the prognosis of HCC patients is still unsatisfactory.One of the most important reasons is that about half of the patients are in the middle or late stage when they were diagnosed and had lost the opportunity of radical surgery.However,there are still no valuable biomarkers that can early diagnosis and predict the prognosis of HCC patients.Therefore,it has become an urgent task for clinicians to explore the molecular mechanism of the tumorigenesis and development of HCC,and find the key molecules for diagnosis and prognosis.According to the previous studies,we selected key genes on HIF and Notch pathways by Bioinformatics analysis,and established diagnostic and prognostic models.Method:Expression data of HCC samples were obtained from the public database TCGA and the datasets GSE25097,GSE36376 of GEO.Genes on the HIF and Notch pathway were obtained by KEGG pathway analysis.The differentially expressed genes(DEGs)were obtained by comparing the expression data of HIF and Notch pathway genes between HCC samples and adjacent tissues in TCGA and GSE25097 datasets.Based on the HCC expression data in TCGA database,the gene network was constructed by Multiscale Embedded Gene Co-expression Network Analysis(MEGENA)to find the genes with similar expression.In addition,according to the survival status and survival time information of patients in TCGA database,the HIF and Notch pathway genes were analyzed to screen out the genes that have a significant impact on the survival of patients.We overlapped the genes obtained from the results of the above analysis,including DEGs,Survival analysis,and MEGENA and selected the genes that overlapped in at least two sets and were consistent with the regulation trend of genes on the TCGA and GSE25097.These genes were used to construct Protein-Protein interaction networks analysis(PPI).Finally,the final candidate genes were obtained by random forest analysis.The diagnostic and prognostic models were constructed based on candidate genes,and the correlation analysis of prognostic characteristics was carried out.The conclusions were verified with the datasets GSE36376.Results:Four candidate genes,including ALDOA,Notch3,MAKP3,SERPINE1,were finally obtained by the above analysis.The ROC curve of the diagnostic model constructed by the candidate genes shows that its sensitivity,specificity and AUC(Area Under The Curve)were all greater than 0.8,indicating that our four candidate genes have a strong ability to identify cancer or not.And the ROC curve of the prognostic model showed that AUC was greater than 0.6,which had a good predictive effect.At the same time,we discussed the predictive ability of the prognostic model in different clinical groups of TCGA HCC,and found that the prognostic model had good predictive effect in different gender,age ≥65,age<65,male,M0,N0,T1-T2,T3-T4 subgroups(the survival rate of the high-risk group was lower than the low-risk group,p<0.05).It had a certain predictive effect in female patients,but there was no statistical significance(p>0.05).Based on the expression and clinical data of HIF and Notch pathway genes in high-risk and low-risk groups,the clinic pathological characteristics were compared by Chi-square test and the results showed that the low-risk group was significantly correlated with lower T stage(p<0.01),lower tumor stage(p<0.01)and higher overall survival rate(p<0.05),while the high-risk group was significantly associated with higher T stage,higher tumor stage and lower overall survival rate,which was consistent with clinical practice.Finally,we analyzed the external datasets GSE36376 and found that the expression up-regulation and down-regulation trends of the four genes in the three data sets were consistent:ALDOA,Notch3 and MAKP3 were significantly up-regulated in HCC and SERPINE1 was significantly down-regulated(all adj p<0.05).Conclusion:ALDOA,Notch3,MAKP3 and SERPINE1 genes may play an important role in the occurrence and development of HCC,and are related to the prognosis of liver cancer.Further study of these four genes will help us understand the molecular mechanism of the occurrence and development of HCC,and provide diagnostic markers and potential therapeutic targets for patients with liver cancer.