Background: Liver hepatocellular carcinoma (LIHC) ranks as the foremost cause of cancer-related deaths worldwide, and its early detection poses considerable challenges. Current prognostic indicators, including alpha-fetoprotein, have notable limitations in their clinical utility, thereby underscoring the necessity for discovering new biomarkers to improve early diagnosis and enable personalized treatment options.
Method: This investigation employed single-cell analysis techniques to identify stem cell-associated genes and assess their prognostic significance for LIHC patients, as well as the efficacy of immunotherapy, utilizing nonnegative matrix factorization (NMF) cluster analysis. A diagnostic model for LIHC was developed and validated through multiple datasets and various machine learning clustering methods. The XGBOOST algorithm identified MRPL17 as the most significant prognostic gene among those associated with stem cells. Additionally, the research explores the relationship between MRPL17 expression and immune cell infiltration. Immunofluorescence staining of LIHC tissue samples was conducted to evaluate the expression and prognostic value of MRPL17, as well as its correlation with KI67.
Results: Through single-cell analysis, this study identified 14 essential stem cell-related genes, highlighting their significance in the diagnosis, prognostication, and potential treatment strategies for LIHC patients. Various machine learning algorithms indicated that MRPL17 is particularly associated with patient prognosis and responses to immunotherapy. Furthermore, experimental results demonstrate that MRPL17 is upregulated in LIHC and correlates with poor prognosis, as well as positively correlating with KI67.
Conclusion: Cancer stem cells are pivotal in the mechanisms of immune evasion within the tumor microenvironment and have a substantial impact on treatment results. This study experimentally validated MRPL17 as a promising prognostic biomarker, emphasizing the need to target liver cancer stem cells to improve patient prognosis and enhance treatment effectiveness.
Keywords: MRPL17; cancer stem cell; hepatocellular carcinoma; machine learning; single cell analysis.
Copyright © 2024 Shao, Zhao, Liu and Kang.