构建基于铁死亡相关基因特征的乳腺癌患者预后预测模型OA

OBJECTIVE:To explore construction of a prognostic prediction model for breast cancer patients based on characteristics of ferroptosis-related genes.METHODS:Functional prediction of breast cancer-related genes was performed using information from the Cancer Genome Atlas Database.Genes associated with ferroptosis were selected,and LASSO regression was used to further construct a prognostic model which was validated for its function and accuracy.Gene set variation analysis(GSVA)was used to comprehensively score each gene set and assess potential changes in biological function among different samples.Patients were divided into high-risk and low-risk groups based on the median risk score of the model,and gene set enrichment analysis(GSEA)was used to further compare differences in signaling pathways between the two groups.A multivariate regression model was constructed,and each influencing factor was scored according to its contribution to the outcome variable(the magnitude of the regression coefficient).The scores were summed to obtain the total score,thereby calculating the predicted value.A weighted gene co-expression network(WGCNA)was constructed to identify co-expressing gene modules,and core genes were obtained by taking the intersection of the model and the modules.Transcription factors were predicted using the R package"RcisTarget".Positional relationships of each gene were obtained through single-cell sequencing,and single nucleotide polymorphism(SNP)sites significantly associated with breast cancer were screened through genome-wide association studies to determine pathogenic regions of key gene SNPs.RESULTS:From 721 samples,67 genes were found to be significantly associated with both ferroptosis and breast cancer prognosis.LASSO regression analysis yielded 20 model genes and the optimal risk score for each sample.Receiver operating characteristic(ROC)curve analysis showed that the model had good predictive power for BRCA patients'survival.Functional enrichment analysis revealed that these genes were mainly enriched in RNA degradation and ferroptosis.GSVA and GSEA analyses showed significantly enriched and differentially expressed pathways between high-risk and low-risk score groups,including glycolysis,unfolded protein response,cell cycle,central carbon metabolism,and MAPK signaling pathways.Regression analysis showed that the risk score of ferroptosis-related genes significantly contributed to the predictive model,demonstrating good consistency in predicting 3-year and 5-year survival rates.Ten gene modules were identified using the constructed WGCNA network and the TOM matrix.A total of 480 genes were extracted from the strongest two modules,and their intersection with the 20 model genes revealed three core genes:VDAC3,PSME1,and CCL5.Single-cell sequencing clarified the enrichment locations of these genes in the cell population.The pathogenic SNP regions of the three core genes were determined:VDAC3 was located on chromosome 8,PSME1 on chromosome 14,and CCL5 on chromosome 17.CONCLUSION:A novel breast cancer prognostic model was constructed based on expression of ferroptosis-related genes and it demonstrated good predictive performance and clinical application potential.