Exploring the significance and potential mechanisms of hippo pathway-associated genes in prognosis of glioma patients

XuKai Liu,Hongjun Fan,Zebo Chen,Chao Liu

DOI: https://doi.org/10.1007/s12672-024-01391-5

2024-10-09

Abstract:Purpose: Despite the efforts of countless researchers to develop glioma treatment strategies, the current therapeutic effect of glioma is still not ideal, and it is necessary to further explore the mechanism to guide treatment. Thus, this study aims to introduce a novel approach for predicting patient prognosis and guiding further treatment interventions. Methods: Initially, we conducted a differential gene expression analysis to identify Hippo pathway-associated genes overexpressed in tumors and determined genes correlated with prognosis. Subsequently, employing cluster analysis, we categorized samples into two groups and performed further analyses including prediction, immune cell infiltration abundance, and drug response rates. We utilized weighted gene co-expression analysis to reveal gene sets with high co-variation, delineate inter-sample gene correlation patterns, and conduct enrichment analysis. Prognostic models were built using ten machine learning algorithms combined in 101 different combinations, followed by evaluation and validation. Immune infiltration analysis, differential expression analysis of depleted T cell-related markers, drug sensitivity analysis, and exploration of pathway dysregulation were performed for different risk groups. Quality control and batch integration were performed, and single-cell data were analyzed using dimensionality reduction clustering algorithms and annotation tools to evaluate the activity of the prognostic model in malignant cells. Results: We conducted data filtering to identify genes overexpressed in tumors, intersecting these genes with Hippo pathway-related genes, identifying 62 genes correlated with prognosis, and performing cluster analysis to divide tumor tissues into two groups. Cluster 2 exhibited a poorer prognosis and demonstrated differences in immune cell infiltration. Utilizing weighted gene co-expression analysis on Cluster 2, we identified gene modules, conducted functional enrichment analysis, and delineated pathways. Employing a combined model based on ten machine learning algorithm combinations, we selected the optimal prognostic model system and validated the model's predictive ability within the dataset. Through immune-related analysis and drug sensitivity analysis, we uncovered differences in immune infiltration and varying sensitivities to chemotherapy drugs. Additionally, the enrichment analysis of gene set revealed discrepancies in upregulation within relevant pathways between the high and low-risk groups. Finally, annotation and evaluation of malignant cells via single-cell analysis showed increased activity of the prognostic model and variations in distribution across different prognostic levels in malignant cells. Conclusion: This study introduces a novel approach utilizing the Hippo pathway and associated genes for glioma prognosis research, demonstrating the potential and significance of this method in evaluating the outcome for patients with glioma. These findings hold substantial clinical significance in guiding therapy and predicting outcomes for individuals diagnosed with glioma, offering significant clinical utility.