The Identification Of Methylation-driven Genes And Prognostic Risk Assessment Model For Colorectal Cancer By Bioinformatics

Objectives: DNA methylation-driven genes is significant in the onset and development of colorectal cancer,but their clinical significance is unclear and urgently need to be investigated.This project aimed to explore the clinical significance of DNA methylation-driven genes and focused on those specific and highly expressed DNA methylationdriven genes of CRC to investigate their biological functions and clinical significance as effective biomarkers.Following establishing the risk model based on these genes by statistical analysis,the signature was compared with other models from published articles and analyzed based on the patients' drug treatment response difference.Methods: To identify DNA methylation-driven genes in this study,the " Methylmix " R package was employed with CRC from the Cancer Genome Atlas(TCGA)database,and the "clusterprofiler" R package was leveraged to perform functional analysis of these genes.Then,according to the overall survival(OS)of the patients,COX regression analysis was used to establish a potential prognostic model which was then compared with the published model.Finally,the model was explored to predict the potential performance of therapies on CRC.Results: By screening the RNA sequencing profiles and corresponding DNA methylation data from the Cancer Genome Atlas(TCGA)database,181 genes were identified as methylation-driven genes from the abnormally expressed genes in CRC.In addition,18 of the methylation-driven genes were shown to be related to overall survival(OS)of patients.Based on five survival-related methylation-driven genes,a risk assessment model was constructed to predict the prognosis of CRC patients.Kaplan-Meier analysis and Harrell's C index indicated that the risk assessment model can clearly distinguish patients in high-risk and low-risk groups(P <0.0001 logarithmic test,HR: 2.034,95% CI: 1.419-2.916,C index: 0.655).Through ROC(the Receiver Operating Characteristic)analysis,the sensitivity and specificity of the signature was evaluated.After verification,the risk score obtained by the risk assessment model can be used as an independent biomarker for poor prognosis of colorectal cancer.In addition,different drug treatment responses were observed between the high-risk and low-risk groups,with a complete remission rate of67.2% in the low-risk arm and 53% in the high-risk arm.Therefore,patients in the highrisk group are more likely to have clinically progressive diseases.Finally,a model composed of five methylation-driven genes(POU4F1,NOVA1,MAGEA1,SLCO4C1,IZUMO2)can be used as an independent indicator to improve clinical prognosis.Our findings will provide effective tools to guide clinical treatment.Conclusions: This study used gene expression profiles and 450 k methylation chips to identify 181 methylation driver genes in colorectal cancer.At the same time,a model composed of POU4F1,NOVA1,MAGEA1,SLCO4C1,and IZUMO2 is obtained according to the effective evaluation indicators,which can effectively provide technical support for patient prognosis and clinical treatment.