Generalized machine learning based on multi‐omics data to profile the effect of ferroptosis pathway on prognosis and immunotherapy response in patients with bladder cancer.

Introduction: Bladder cancer (BLCA) affects millions of people worldwide, with high rates of incidence and mortality. Ferroptosis proves to be a novel form of cell death process that is triggered by oxidative stress. Methods: We procured a total of 25 single nuclear RNA‐seq (snRNA‐seq) samples from GSE169379 in GEO database. We obtained different cohorts of BLCA patients from the TCGA and GEO databases for model training and validation. A total of 369 ferroptosis‐related genes (FRGs) were selected from the FerrDb database. AUCell analysis was performed to assign ferroptosis scores to all the cell types. Weighted Gene Co‐Expression Network Analysis (WGCNA), COX, and LASSO regression analysis were conducted to retain and finalize the genes of prognostic values. Various bioinformatic approaches were utilized to depict immune infiltration profile. We conducted a series of colony formation analysis, flow cytometry and western blot (WB) analysis to determine the role of SKAP1 in BLCA. Results: We divided the cells into high ferroptosis group and low ferroptosis group according to ferroptosis activity score, and then screened 2150 genes most associated with ferroptosis by differential expression analysis, which are related to UV‐induced DNA damage, male hormone response, fatty acid metabolism and hypoxia. Subsequently, WGCNA algorithm further screened 741 ferroptosis related genes from the 2150 genes for the construction of prognostic model. Lasso‐Cox regression analysis was used to construct the prognostic model, and the prognostic model consisting of 6 genes was obtained, namely JUN, SYT1, MAP3K8, GALNT14, TCIRG1, and SKAP1. Next, we constructed a nomogram model that integrated clinical factors to improving the accuracy. In addition, we performed drug sensitivity analyses in different subgroups and found that Staurosporine, Rapamycin, Gemcitabine, and BI‐2536 may be candidates for the drugs treatment in high‐risk populations. The ESTIMATE results showed higher stromal scores, immune scores, and ESTIMATE scores in the low‐risk group, indicating a higher overall immunity level and immunogenicity of tumor microenvironment (TME) in this group, and tumor immune dysfunction and exclusion (TIDE) analysis confirmed a better response to immunotherapy in the low‐risk group. Finally, we selected the oncogene SKAP1 in the prognostic gene for in vitro validation, and found that SKAP1 directly regulated BLCA cell proliferation and apoptosis. Conclusion: We identified a set of six genes, JUN, SYT1, MAP3K8, GALNT14, TCIRG1, and SKAP1, that exhibited significant potential in stratification of BLCA patients with varying prognosis. In addition, we uncovered the direct regulatory effect of SKAP1 on BLCA cell proliferation and apoptosis, shedding some light on the role of FRGs in pathogenesis of BLCA. [ABSTRACT FROM AUTHOR]