Integrating genetic and gene expression data in network-based stratification analysis of cancers

Area of Science:

• Oncology
• Computational Biology
• Genomics

Background:

• Cancer is a complex disease with diverse genetic drivers and outcomes.
• Effective prognosis and treatment require organizing patient cohorts into biologically and clinically relevant subtypes.
• Omics data offers a rich resource for cancer patient stratification, but integrating these diverse data types remains a challenge.

Purpose of the Study:

• To investigate a novel approach to network-based stratification (NBS) by integrating somatic mutation and RNA sequencing data.
• To evaluate the effectiveness of integrated NBS for cancer subtyping in ovarian, bladder, and uterine cancers.
• To identify clinically and biologically meaningful cancer subtypes using multi-omics data.

Main Methods:

• Applied network-based stratification (NBS) framework.
• Integrated somatic mutation data with RNA sequencing data for three cancer types.
• Compared integrated NBS subtypes with single-data type NBS subtypes for associations with patient survival and tumor histology.

Main Results:

• Integrated NBS subtypes showed significantly stronger associations with overall survival and histology compared to single-data type NBS subtypes.
• Specific improvements were observed for ovarian and bladder cancer survival associations, and for bladder and uterine cancer histology associations.
• Identified influential genes and pathways (e.g., ubiquitin homeostasis, p53 regulation, cytokine signaling) underlying integrated NBS subtypes, revealing biological differences.

Conclusions:

• Integrating multi-omics data within the NBS framework significantly enhances cancer subtyping.
• This approach has profound implications for personalized prognosis and treatment strategies.
• The findings advance computational subtyping methods and facilitate the discovery of cancer driver genes.