Bioinformatics Analysis Of Colorectal Cancer Related Genes And Functional Prediction Of ABCD3

Background: Colorectal cancer is one of the most common malignancies of the digestive system With multiple causes and poor prognosis,new efficient gene targets and molecular markers need to be further studied.Recently,bioinformatics mining has been widely used to find potential targets of malignancies.It has been reported that ABCD3(ATP binding cassette subfamily D member 3)had diagnostic and prognostic value in many tumors,but there are few studies about colorectal cancer.Objective: To learn about biological process pathways and to explore new molecular markers of colorectal cancer.To study ABCD3 expression in colorectal cancer and to explore diagnostic and prognostic value of ABCD3.Methods: The microarray datasets from GEO was analyzed by R language,DAVID6.8,KOBAS2.0 and STRING10,and the differential genes in colorectal cancer were screened out.Then GO analysis,KEGG analysis and protein-protein interaction(PPI)network were constructed.R language,survival analysis was performed with TCGA colorectal cancer data.Then SPSS 19.0and GraphPad Prism 6.0 were used for analysis.ABCD3 expression among colon cancer,rectal cancer,all stages of colorectal cancer and normal tissues were compared.Then receiver operating characteristic(ROC)curves and the area of curve(AUC)were performed.Survival analysis and COX regression analysis were made to evaluate the prognostic value of ABCD3.Pearson linear correlation was used to analyzed the relationship between ABCD3 expression and DNA methylation,while Analysis of Variance was used to evaluate relationship between ABCD3 expression and copy number variation.Results: There were 287 differential genes in this test,among which 191 genes were up-regulated and 96 genes were down-regulated.GO analysis showed that the differential genes were enriched in the proteinaceous extracellular matrix(P=2.87E-05,FDR=0.0365)and the extracellular region(P=3.35E-05,FDR=0.0426).KEGG analysis showed there were 26 pathways involved(P< 0.01).PPI showed that central codes included CXCL8,CXCL1,PYY,PPBP,P2RY14,INSL5,CXCL3,CXCL10,CXCL11 and NUF2.There were 331 genes associated with the prognosis including ABCD3.High expression of 223 genes were related tobetter overall survival,while 108 genes were related to worse overall survival.ABCD3 expression in colon cancer,rectal cancer and all stages of colorectal cancer was significantly lower than that in normal tissues.ABCD3 had high diagnostic efficiency in colon cancer(AUC=0.9552,P<0.01),rectal cancer(AUC=0.9713,P<0.01)and stage I-IV colorectal cancer(total AUC=0.9589,ACU in I-IV tumors: 0.9456,0.9578,0.9613,0.9639,P<0.01).ABCD3 expression was negatively correlated with DNA methylation(Pearson’s r =-0.1949,P<0.0001(data from UCSC),Pearson’s r =-0.295,P<0.01(data from cBioPortal)).ABCD3 expression was lower in the copy deletion group and higher in the amplification group(P<0.05).Survival analysis and COX regression analysis showed that patients with low ABCD3 expression had a poor overall survival(P<0.01).Conclusion: Many genes and pathways may play important roles in colorectal cancer.Then a series of genes associated with prognosis were screened,including ABCD3.ABCD3 may have diagnostic and prognostic value in colorectal cancer.DNA methylation and copy number variation may be associated with ABCD3 expression.