Construction Of Immune-related Prognostic Models For Colorectal Cancer And Screening Of Small Molecule Drugs

Objective: Colorectal cancer is currently one of the leading causes of death from malignant tumors worldwide.Immune related genes(IRGs)play an important role in the development of colorectal cancer and are related to the prognosis of patients.Based on bioinformatics,this study established a prognostic model of immune-related genes in colorectal cancer to evaluate the prognosis of colorectal cancer patients.The small molecule drugs screened also provide a potential therapeutic drug for new drug research against colorectal cancer.Methods: 1.The gene expression profiles of 203 colorectal cancer tissues and 160 adjacent tissues were downloaded from the GEO database,the R language "Limma" package was used to screen for differentially expressed genes,the immune genes were obtained from the Imm Port and Innatedb databases,and then the intersection treatment was performed to obtain colorectal cancer immune differential genes.2.The "cluster Profiler" package in R was used to perform GO enrichment analysis and KEGG enrichment analysis of colorectal cancer immune difference genes.3."WGCNA" R was used to construct a co-expression network for colorectal cancer immune difference genes,and the module gene with the strongest correlation with colorectal cancer survival was found.4.The target module genes in WGCNA were analyzed for univariate COX survival,random survival forest analysis and LASSO regression analysis,and finally the immunerelated prognostic model genes and their risk factors were obtained.5.According to the median risk factor,the prognostic model was divided into high and low risk groups,and colorectal cancer data was downloaded from TCGA database as a test group to verify the reliability of the prognostic model,and the time-dependent ROC curve and nomogram were used to verify the model.6.The genes with statistical significance in univariate COX survival analysis were divided into up-regulated group and down-regulated group,and imported into the Cmap database in order to explore small molecule drugs that may treat colorectal cancer,and a negative score indicates that the drug reverses the expected biological characteristics and has potential therapeutic value.7.Molecular docking verification: verify the binding of small molecule drugs to target genes through molecular docking software.Results: 1.Through the mining of differential genes between colorectal cancer tissues and adjacent tissues,a total of 2896 differential genes were identified,and after intersecting with the immune genes obtained in the database search,479 immune-related differential genes were finally screened,including 232 upregulated genes and 247 downregulated genes.2.GO pathway analysis showed that immune differential genes were mainly enriched in gene coding products and were involved in biological processes such as cytokinemediated signaling,leukocyte migration,and active regulation of external stimuli responses.KEGG pathway analysis identified 15 pathways associated with tumor-related genes,including cytokine-cytokine receptor interactions,chemokine signaling pathways,and complement and coagulation cascades in cancer.3.Through the construction of WGCNA co-expression network,gene modules(cor=0.87,P=7e-114)that were closely related to the survival time of colorectal cancer patients were screened,with a total of 142 genes in them.4.Univariate COX survival analysis was used to evaluate the prognostic correlation between each gene and overall survival(OS)of patients,and found that 42 genes had statistical differences(P<0.05),of which 14 genes were protective factors(HR<1)and 28 genes were risk factors(HR>1).5.Through random survival forest analysis and LASSO regression analysis,nine immune-related prognostic model genes and their risk factors were obtained,namely A2 M,BCL2L2,CCR2,CD2 AP,CMA1,IL6 ST,MLLT3,NEDD4 and SCG2.Through the independent prognostic analysis of risk score and prediction nomogram,the predictability of the model is proved.6.Nine genes were validated using TCGA data,and the AUC values for 1,3,and 5years in the GEO cohort were 0 93,0.82,0.82;The AUC values for years 1,3,and 5 in the TCGA queue are 0.71,0.70,0.78.The independent prognostic analysis and predictive column chart of risk scores demonstrate the predictive ability of the model.7.The six potential small molecule drugs with the highest negative normalized connection scores were screened out through the Cmap database.They are prothionamide,granisetron,OMDM-2,hypericin,imperatorin and fluvastatin.8.Molecular docking verifies the binding pattern of small molecule drugs to target genes,and Hypericin is considered a potential therapeutic agent for colorectal cancer.Conclusion: The prognosis model of immune-related genes in colorectal cancer constructed in this study can effectively predict the prognosis of patients,and in addition,the screened small molecule drug Hypericin provides a new idea for the development of colorectal cancer drugs.