Application of Kaplan-Meier Estimator Model for Validation of AURKC as an Early Biomarker for Kidney Cancer

Abstract

Artificial neural networks are very powerful algorithms for predicting cancer survival results but evaluating their accuracy is vital and required for successful clinical application. This research introduces an enhanced Kaplan -Meier estimator model that contains new additional features for better validating AURKC as an early-stage prediction marker in kidney cancer. The enhancements include using a larger sample size (530 kidney cancer cases) and integrating additional variables such as AURKC expression, disease stage -1, follow-up threshold up to 150 months, sex, and race. These enhancements focus on improving predictive accuracy. Using public datasets TCGA, EGA, and GEO, genes linked to survival changes outcomes were detected. Log rank regression revealed AURKC as the top prognostic gene with a hazard rate of 2.36. High levels of AURKC linked to shorter survival in stage-1 were found in white male patients. In conclusion, integration of multi database analysis and advanced statistical models validates the identification and prioritizing of AURKC as a promising biomarker and as a target for drug development in malignancies, as well as kidney cancer.