Research Paper Volume 16, Issue 4 pp 3332—3349
Machine learning constructs a T cell-related signature for predicting prognosis and drug sensitivity in ovarian cancer
- 1 Department of Obstetrics and Gynecology, General Hospital of Northern Theater Command, Shenyang 110015, China
Received: July 20, 2023 Accepted: December 7, 2023 Published: February 9, 2024
https://doi.org/10.18632/aging.205536How to Cite
Copyright: © 2024 Zhang and Pei. This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Abstract
Background: The leading cause of death related to gynecologic cancer is ovarian cancer, which typically has a poor prognosis. T cells are referred to as key mediators of immunosurveillance and tumor eradication, and unbalanced regulation or lack of T cells in tumors result in immunotherapy resistance.
Methods: The identification of T cell related markers depended on single-cell RNA-seq analysis. Using data from multiple datasets, including TCGA, GSE14764, GSE26193, GSE26712, and GSE140082, we constructed a prognostic signature called TRS (T cell-related signature) using 10 different machine learning algorithms. The correlation between TRS and drug sensitivity were analyzed using the data from GSE91061 and IMvigor210 dataset.
Results: PlsRcox method based TRS was as a risk factor for the clinical outcome of ovarian cancer patients. In comparison with stage, grade and many prognostic signatures, the performance of our TRS in evaluating the clinical outcome was better in ovarian cancer. TRS-based risk score showed distinct association with the level of ESTIMATE score, immune-related function score and immune cells. Moreover, TRS could be used to predict the immunotherapy response and chemotherapy response in ovarian cancer.
Conclusion: In conclusion, we constructed a powerful TRS in ovarian cancer, which could accurately predict the clinical outcome of patients and be used to predict the immunotherapy response and chemotherapy response.