Related Experiment Video
Updated: Jan 12, 2026
08:25
Mass Cytometry Analysis of Systemic and Local Immune Responses in Hepatocellular Carcinoma
Published on: April 25, 2025
699
Comprehensive multi-omics data to construct hepatocellular carcinoma pathway subtypes and classification model
Lurao Ji1, Xiaoqin Li1, Bin Gao1
1College of Chemistry and Life Science of Beijing University of Technology, Beijing 100124, China.
Computational Biology and Chemistry
|November 1, 2025
Summary
This study developed a new pathway-based classification for Hepatocellular Carcinoma (HCC) using multi-omics data, identifying three distinct subtypes with varying prognoses and potential drug sensitivities. The research also uncovered novel telomere-related biomarkers for personalized HCC treatment.
Area of Science:
- Oncology
- Bioinformatics
- Genomics
Background:
- Hepatocellular Carcinoma (HCC) exhibits high heterogeneity, complicating clinical outcomes and treatment strategies.
- Existing gene expression-based classifications may not fully represent the complex biological mechanisms and phenotypic changes in HCC patients.
Purpose of the Study:
- To construct novel HCC molecular subtypes by integrating multi-omics data and pathway information.
- To identify subtype-specific therapeutic targets and prognostic biomarkers.
- To develop a robust and reproducible classification model for HCC.
Main Methods:
- Integration of multi-omics data (gene expression, mutation) with gene set pathway analysis (GSVA).
- Construction of three HCC subtypes (PS1, PS2, PS3) based on pathway activity and genomic data.
- Identification of subtype-specific drugs, core pathways, telomere-related biomarkers, and a prognostic nomogram.
- Development of a machine learning-based classification model using 10 genes.
Main Results:
- Three distinct HCC subtypes (PS1, PS2, PS3) were identified, showing significant differences in immune infiltration, pathway characteristics, and genomic instability.
- PS3 subtype exhibited the worst prognosis, with enrichment in cell proliferation pathways.
- Six subtype-specific drugs were identified, with PS1 showing higher sensitivity to KU_55933 and Cyclophosphamide.
- Three telomere-associated biomarkers (POLD1, RFC1, TERF1) were identified.
- A reproducible classification model (NN_MLP_10) with high accuracy (AUC = 0.930) was established.
Conclusions:
- The study successfully established a pathway-based molecular subtyping system for HCC, revealing significant multi-omics differences between subtypes.
- The identified telomere-associated biomarkers and subtype-specific drugs provide a basis for personalized treatment strategies in HCC.
- The developed classification model offers a reproducible tool for clinical application and patient stratification.
