An efficient strategy for identifying cancer-related key genes based on graph entropy

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Gene networks are crucial for identifying clinically relevant functional genes.
• Existing methods often rely on gene/protein interaction data, which can be incomplete, leading to biased conclusions.
• The human interactome's incompleteness poses a challenge for constructing accurate gene networks.

Purpose of the Study:

• To propose an efficient strategy for identifying cancer-related key genes using gene expression and mutation data.
• To overcome the limitations of incomplete human interactome data in gene network construction.
• To develop a novel metric combining graph entropy and gene mutation influence for assessing gene impact.

Main Methods:

• Constructing a gene network using gene expression data via a sparse inverse covariance matrix.
• Clustering genes using parallel maximal cliques algorithm to obtain subgraphs.
• Developing a novel metric integrating graph entropy and upstream gene mutation influence to measure gene impact factors.

Main Results:

• The proposed strategy effectively extracts key genes involved in tumorigenesis from cancer datasets.
• The identified key genes demonstrate distinct roles in cancer development.
• Cancer patient risk groups are accurately predicted using the identified key genes.

Conclusions:

• The iKGGE strategy provides an efficient and robust method for identifying cancer-related key genes.
• This approach enhances the reliability of gene network analysis by mitigating reliance on incomplete interactome data.
• The identified key genes hold potential for improving cancer diagnostics and risk stratification.