Construction And Validation Of An Immune Landscape Related Gene Signature From Colorectal Cancer By Bioinformatics

Objective:Colorectal Cancer(CRC)is the second leading cause of cancer-related deaths in our country.Tumor Microenvironment(TME)refers to the cellular environment where tumor and cancerous stem cells survive,which plays a crucial role in mediating tumor immunity and tumor progression.Single sample Gene Set Enrichment Analysis(ssGSEA)can obtain the abundance of immune cell infiltration and immune-related function in TME(immune landscape).Therefore,this study intends to discover the relationship between tumor immune landscape and CRC through ssGSEA and establish a reliable prognostic model.Method:Data were collected from the Tumor Genome Atlas(TCGA)and the Gene Expression Dataset(GEO).TCGA transcriptome data were used for ssGSEA to obtain sample immunologic scores,and then CRC patients were divided into high and low immunity subgroups using unsupervised hierarchical clustering.The ESTIMATE algorithm was used to calculate stroma/immune cell scores and their relationship to clinical characteristics in CRC patients.Screening the differentially expressed genes of high and low immune subgroups and intersecting with immune-related genes,Univariate and multivariate Cox regression analysis was used to construct the Tumor immune Microenvironment related score(TIMERS)model with statistically different genes by Cox analysis.The samples’ risk values were calculated and the high and low risk subgroups were divided according to the median risk value.Survival analysis showed statistically significant differences.To further verify the robustness of the model,GEO data set was used for external validation.Independent prognostic analysis was used to determine whether TIMERS was an independent prognostic factor for CRC.Subsequently,a nomogram combining TIMERS and clinical risk factors was constructed to predict the overall survival of CRC patients at 1,3 and 5 years,and the accuracy of the nomogram was verified by time-dependent ROC curve and calibration curve.Finally,R software pRRophetic package and TCI A website were used to analyze the response of CRC patients in the high and low risk subgroups to predict chemotherapy and immunotherapy efficacy.Results:In this study,the data of 568 CRC cases in the TCGA database were were used to ssGSEA and hierarchical clustering,and the tumor immune landscape differences between the high and low immune groups was analyzed.Immune-related genes were downloaded from the IMMPORT database,and the Venn diagram were used to describe the intersection of differentially expressed genes and immune-related genes in the high and low immune subgroups(344 genes).After univariate and multivariate Cox regression analysis,a prognostic model containing CX3CL1,NOS2,APOBEC3F,CCL22 and ANGPTL4 was established.It has been proved that the model could well distinguish high and low risk patients and predict survival.Independent prognostic analysis showed that the TIMERS was an independent prognostic factor for CRC.The nomogram constructed by combining TIMERS and clinical risk factors can well predict the overall survival of CRC patients at 1,3 and 5 years(AUC values at 1,3 and 5 years were 0.771,0.784 and 0.802,respectively).Src inhibitors may be more sensitive in high-risk patients,and analysis by TCIA website suggests that low-risk patients are more sensitive to immunotherapy.Conclusions:1.In this study,five prognostic related genes were analyzed and a prognostic risk prediction model based on CX3CL1,NOS2,APOBEC3F,CCL22 and ANGPTL4 was constructed,and the robustness of the model was verified.2.The nomogram constructed by TIMERS and clinical risk factors can accurately predict the overall survival rate of patients at 1,3 and 5 years.3.Src inhibitors may be more sensitive in TIMERS high-risk patients and more sensitive to immunotherapy in TIMERS low-risk patients.This study may help to stratify CRC risk and guide clinical practice.