Wei Tan

Pan‐cancer analysis of oncogenic role of insulin‐like growth factor‐binding proteins and validation in ovarian cancer

AbstractBackgroundNumerous studies have shown that the insulin‐like growth factor (IGF) pathway is highly associated with tumor initial and progression in several tumors. However, compared with IGF1/1R and IGF2/2R, insufficient studies have focused on IGF‐binding proteins (IGFBPs).MethodsThe GDC TCGA and GTEx data of 33 cancers, TCGA pan‐cancer immune phenotypes, tumor mutation burdens, and the copy number alterations of IGFBPs were extracted. Next, the prognostic value of IGFBPs was analyzed based on a univariate Cox analysis. Additionally, the ESTIMATE algorithm was used to calculate stromal and immune scores and tumor purity, and the CIBERSORT algorithm was used to estimate tumor‐infiltrating immunocyte levels. Ultimately, the correlation between IGFBP expression and cancer hallmark pathways was estimated with a Spearman analysis.ResultsThe expression of IGFBPs was differentially expressed and correlated with prognosis in specific cancers. IGFBPs may operate as biological markers for carcinogenesis and progression and as prognostic biomarkers. Additionally, IGFBP5 has been proved that promotes the invasion and migration of ovarian cancer.ConclusionsIn general, IGFBPs can serve as predictable biomarkers and potential therapeutic targets for specific tumors. Our results could provide underlying targets for the design of laboratory experiments to elucidate the mechanism of IGFBPs in cancers and identify IGFBP5 as a prognostic factor in ovarian cancers.

Gene signature of m6A-related targets to predict prognosis and immunotherapy response in ovarian cancer

The aim of the study was to construct a risk score model based on m6A-related targets to predict overall survival and immunotherapy response in ovarian cancer. The gene expression profiles of 24 m6A regulators were extracted. Survival analysis screened 9 prognostic m6A regulators. Next, consensus clustering analysis was applied to identify clusters of ovarian cancer patients. Furthermore, 47 phenotype-related differentially expressed genes, strongly correlated with 9 prognostic m6A regulators, were screened and subjected to univariate and the least absolute shrinkage and selection operator (LASSO) Cox regression. Ultimately, a nomogram was constructed which presented a strong ability to predict overall survival in ovarian cancer. CBLL1, FTO, HNRNPC, METTL3, METTL14, WTAP, ZC3H13, RBM15B and YTHDC2 were associated with worse overall survival (OS) in ovarian cancer. Three m6A clusters were identified, which were highly consistent with the three immune phenotypes. What is more, a risk model based on seven m6A-related targets was constructed with distinct prognosis. In addition, the low-risk group is the best candidate population for immunotherapy. We comprehensively analyzed the m6A modification landscape of ovarian cancer and detected seven m6A-related targets as an independent prognostic biomarker for predicting survival. Furthermore, we divided patients into high- and low-risk groups with distinct prognosis and select the optimum population which may benefit from immunotherapy and constructed a nomogram to precisely predict ovarian cancer patients' survival time and visualize the prediction results.