Pancreatic cancer (PC) has a poor prognosis, and treatment is largely ineffective because of its unique tumor microenvironment (TME). We identified key cell types in PC using the GSE197177 dataset generated through single-cell RNA sequencing (scRNA-seq) analysis. Using TCGA-PAAD data, prognosis-related cell subtypes were identified by BayesPrism and Cox regression. Hub cell subtype-related prognostic genes were screened by integrating high-dimensional weighted gene co-expression network analysis (hdWGCNA), differential expression analysis, univariate Cox regression, and least absolute shrinkage and selection operator (LASSO) regression. The upstream regulatory factors of prognostic genes were predicted. A risk model and nomogram were generated and validated, with risk scores used to evaluate pathways, the TME, immunotherapy, and drug sensitivity. Fibroblasts were identified as the key cell type in PC. TPM1+ myofibroblastic cancer-associated fibroblasts (CAFs) (myCAFs) were considered the prognosis-related hub cell subtype. AHNAK2, ARHGAP32, EVL, PRKCI, PTGES, and S100A16 were identified as prognostic genes. Regulatory factors, including DPF2 and hsa-miR-107, were predicted to target the prognostic genes. The constructed risk model and nomogram showed promise in prognostic performance, although further validation in diverse clinical subgroups is needed. Risk scores were associated with pathways, including the cell cycle, infiltration of immune cell types such as T follicular helper cells, response to immune checkpoint blockade therapy, and sensitivity to drugs such as lapatinib. A risk model based on six TPM1+ myCAF-related prognostic genes was constructed, which exhibited robust predictive ability. Our results provide novel insights into TPM1+ myCAF-related mechanisms and PC prognostic prediction, although the model’s generalizability to all patient subgroups requires further examination.
更多