Machine learning and multi-omics data reveal driver gene-based molecular subtypes in hepatocellular carcinoma for...

Area of Science:

• Oncology
• Genomics
• Bioinformatics

Background:

• Hepatocellular Carcinoma (HCC) exhibits significant heterogeneity, hindering effective treatment strategies.
• Molecular subtyping of HCC is essential for developing personalized anti-tumor therapies.
• The role of driver genes in HCC subtyping remains underexplored.

Purpose of the Study:

• To leverage driver genes for constructing Hepatocellular Carcinoma subtype models.
• To elucidate the molecular mechanisms underlying identified HCC subtypes.
• To develop a robust prognostic model for HCC patient stratification.

Main Methods:

• Expanded a set of driver genes using computational frameworks based on mutation and dysregulation data.
• Developed a multi-omics subtype classification algorithm integrating mutation and expression data.
• Utilized single-cell and multi-omics data for in-depth subtype characterization.

Main Results:

• Successfully classified Hepatocellular Carcinoma into two distinct subtypes, CLASS A and CLASS B, with significant survival differences.
• Revealed substantial transcriptomic, mutational, copy number variation, and epigenomic distinctions between subtypes.
• Developed a 10-gene classification model, identifying TTK as a potential therapeutic target.

Conclusions:

• Driver gene-based subtyping offers a novel approach to understanding Hepatocellular Carcinoma heterogeneity.
• The identified subtypes and prognostic model provide crucial insights for HCC pathogenesis and treatment development.
• TTK emerges as a promising therapeutic target for specific HCC subtypes.