Construction and validation of a prognostic model for predicting clear cell renal cell carcinoma based on complement-related genes

Clear cell renal cell carcinoma (ccRCC) is a highly heterogeneous tumor and is the most common subtype of renal cell carcinoma (RCC). Surgery is used to cure most early ccRCC, but the 5-year overall survival (OS) of ccRCC patients is far from satisfactory. Thus, new prognostic features and therapeutic targets for ccRCC need to be identified. Since complement factors can influence tumor development, we aimed to develop a model to predict the prognosis of ccRCC through complement-related genes.Differentially expressed genes were screened from an International Cancer Genome Consortium (ICGC) data set, and the genes associated with prognosis were screened by univariate regression and least absolute shrinkage and selection operator-Cox regression, and column line plots were generated using the rms R package to predict OS. The C-index was used to show the accuracy of the survival prediction and the prediction effects were verified using a data set from The Cancer Genome Atlas (TCGA). An immuno-infiltration analysis was performed with CIBERSORT analysis, and a drug sensitivity analysis was performed using the Gene Set Cancer Analysis (GSCA) (http://bioinfo.life.hust.edu.cn/GSCA/#/) database.We identified 5 complement-related genes (i.e., A2M, APOBEC3G, COL4A2, DOCK4, and NOTCH4) for risk-score modeling to predict OS at 1, 2, 3, and 5 years, and the C-index of the prediction mode was 0.795. In addition, the model was successfully validated in TCGA data set. The CIBERSORT analysis showed that M1 macrophages were downregulated in the high-risk group. The GSCA database analysis showed that DOCK4, COL4A2, and A2M were positively correlated with the half maximal inhibitory concentration (IC50) of 10 drugs and small molecules, and COL4A2, NOTCH4, A2M, and APOBEC3G were negatively correlated with the IC50 of dozens of different drugs and small molecules.We developed and validated a survival prognostic model based on 5 complement-related genes for ccRCC. We also elucidated the relationship with tumor immune status and developed a new predictive tool for clinical purposes. In addition, our results showed that A2M, APOBEC3G, COL4A2, DOCK4, and NOTCH4 may be potential targets for the treatment of ccRCC in the future.