Integrated transcriptome analysis and combinatorial machine learning to construct a homeostatic model of acetylation for ccRCC and validate the key gene GCNT4

TL;DR

This study uses transcriptome data and machine learning to build a model of acetylation in kidney cancer and identifies GCNT4 as a key gene linked to better patient outcomes.

Contribution

A novel acetylation homeostasis model for ccRCC is constructed using integrated transcriptome analysis and combinatorial machine learning, validating GCNT4 as a key gene.

Findings

84 acetylation-regulated genes were identified, with significant expression differences between tumor and normal tissues.

The LASSO + RSF model accurately predicted patient prognosis, with high-risk patients showing worse survival.

GCNT4 overexpression inhibited cancer cell proliferation and migration, possibly by regulating O-GlcNAc modification levels.

Abstract

Clear cell renal cell carcinoma (ccRCC) is one of the most common malignant tumors of the urinary system. Protein acetylation plays a key role in regulating cellular processes and cancer signaling pathways. This study explores the potential biological mechanisms of ccRCC from the perspective of acetylation.

This study obtained RNA-seq data and clinical information of ccRCC from TCGA and ICGC, and single-cell RNA sequencing datasets from the GEO database. Ten machine learning algorithms and their 101 combinations were used to analyze the prognostic significance of acetylation-related differentially expressed genes (DEGs) and to construct a prognostic risk model. GSEA was used to analyze the enrichment of different signaling pathways in high-risk and low-risk groups, and the correlation between immune infiltration and risk scores was assessed. Finally, the function of the key gene GCNT4 was verified through cell experiments.

This study identified 84 acetylation-regulated key genes with significant expression differences between tumor and normal tissues, closely linked to patient prognosis. The LASSO + RSF combination model performed best, and the model could accurately predict patient prognosis. The survival of patients in the high-risk group was significantly worse than that in the low-risk group. High expression of GCNT4 was associated with better survival prognosis and was expressed at higher levels in normal tissues than tumor tissues. Overexpression of GCNT4 significantly inhibited the proliferation, invasion, and migration of renal cancer cells and may affect acetylation by regulating the levels of O-GlcNAc modification in cells.

This study constructed a ccRCC acetylation homeostasis model via transcriptome analysis and machine learning, validating GCNT4 as a key gene. High expression of GCNT4 is associated with better survival prognosis and affects acetylation by regulating O-GlcNAc modification levels, inhibiting the proliferation and migration of renal cancer cells, providing a new potential target for the treatment of ccRCC.

The online version contains supplementary material available at 10.1186/s12935-025-03837-4.

Figures