A Cancer Gene Module Mining Method Based on Bio-Network of Multi-Omics Gene Groups

Area of Science:

• Genomics
• Bioinformatics
• Cancer Research

Background:

• Cancer development is influenced by environmental and genetic factors.
• High-throughput sequencing has advanced the understanding of oncogenes and their interactions.
• Research is ongoing to uncover pathogenic mechanisms driving cancer.

Purpose of the Study:

• To integrate heterogeneous multi-level data, including long non-coding RNA (lncRNA) omics data.
• To construct multi-omics bio-network models for screening key cancer-related gene groups.
• To develop a novel clustering algorithm for excavating key gene modules involved in abnormal regulation.

Main Methods:

• Studied integrated multi-level data, incorporating lncRNA omics data.
• Developed a compactness clustering algorithm based on corrected cumulative rank scores.
• Used functional similarity between gene groups as a distance measure for clustering.
• Performed survival analysis on the identified gene groups and modules.

Main Results:

• Successfully constructed multi-omics bio-network models.
• Screened key cancer-related gene groups and identified key gene modules.
• The proposed clustering algorithm effectively excavated abnormally regulated gene groups.
• Survival analysis demonstrated the model's ability to stratify groups effectively.

Conclusions:

• Integration of multi-omics data enables the discovery of key gene modules and dysregulated gene groups in cancer.
• The developed method is crucial for advancing cancer research by identifying critical regulatory elements.
• This approach provides valuable insights into the complex mechanisms underlying cancer occurrence and progression.