| Summary: | This study sought to perform integrative analysis of the immune/methylation/autophagy landscape on breast cancer prognosis and single-cell genotypes. Breast Cancer Recurrence Risk Score (BCRRS) and Breast Cancer Prognostic Risk Score (BCPRS) were determined based on 6 prognostic IMAAGs obtained from the TCGA-BRCA cohort. BCRRS and BCPRS, respectively, were used to construct a risk prediction model of overall survival and progression-free survival. Predictive capacity of the model was evaluated using clinical data. Analysis showed that BCRRS is associated with a high risk of stroke. In addition, PPI and drug-ceRNA networks based on differences in BCPRS were constructed. Single cells were genotyped through integrated scRNA-seq of the TNBC samples based on clustering results of BCPRS-related genes. The findings of this study show the potential regulatory effects of IMAAGs on breast cancer tumor microenvironment. High AUCs of 0.856 and 0.842 were obtained for the OS and PFS prognostic models, respectively. scRNA-seq analysis showed high expression levels of adipocytes and adipose tissue macrophages (ATMs) in high BCPRS clusters. Moreover, analysis of ligand-receptor interactions and potential regulatory mechanisms were performed. The LINC00276&MALAT1/miR-206/FZD4-Wnt7b pathway was also identified which may be useful in future research on targets against breast cancer metastasis and recurrence. Neural network-based deep learning models using BCPRS-related genes showed that these genes can be used to map the tumor microenvironment. In summary, analysis of IMAAGs, BCPRS, and BCRRS provides information on the breast cancer microenvironment at both the macro- and microlevels and provides a basis for development of personalized treatment therapy.
|
|---|
