Introduction: Endometrial cancer is currently one of the three most common female reproductive cancers, which seriously threatens women’s lives and health. Hypoxia disrupts the tumor microenvironment, thereby affecting tumor progression and drug resistance.Methods: We established hypoxia-related gene model to predict patient prognosis and 1-, 3-, and 5-year overall survival rates. Then, the expression level of hypoxia-related genes and survival data were extracted for comprehensive analysis by Cox regression analysis, and the model was established.Results: We analyzed the survival and prognosis of patients in the high and low-risk groups. The Kaplan-Meier curve showed that the low-risk group is associated with a better survival rate. The 1-, 3-, and 5-year AUC values of the model were 0.680, 0.698, and 0.687, respectively. Finally, we found that LAG3 may be a potential immune checkpoint for endometrial cancer.Conclusion: We found four hypoxia-related genes (ANXA2, AKAP12, NR3C1, and GPI) associated with prognosis. The hypoxia-related gene model can also predict prognosis and tumor microenvironment in endometrial cancer.

Early diagnosis is critical for improving outcomes in cancer patients; however, the application of diagnostic markers derived from serum proteomic screening remains challenging. Artificial intelligence (AI), encompassing deep learning and machine learning (ML), has gained increasing prominence across various scientific disciplines. In this study, we utilized cervical cancer (CC) as a model to develop an AI-driven pipeline for the identification and validation of serum biomarkers for early cancer diagnosis, leveraging mass spectrometry-based proteomics data. By processing and normalizing serum polypeptide differential peaks from 240 patients, we employed eight distinct ML algorithms to classify and analyze these differential polypeptide peaks, subsequently constructing receiver operating characteristic (ROC) curves and confusion matrices. Key performance metrics, including accuracy, precision, recall, and F1 score, were systematically evaluated. Furthermore, by integrating feature importance values, Shapley values, and local interpretable model-agnostic explanation (LIME) values, we demonstrated that the diagnostic area under the curve (AUC) achieved by our multi-dimensional learning models approached 1, significantly outperforming the diagnostic AUC of single markers derived from the PRIDE database. These findings underscore the potential of proteomics-driven integrated machine learning as a robust strategy to enhance early cancer diagnosis, offering a promising avenue for clinical translation.