Integrating multi-omics data to identify dysregulated modules in endometrial cancer

Area of Science:

• Genomics
• Bioinformatics
• Cancer Research

Background:

• Cancer arises from genetic mutations, with differentially expressed genes closely linked.
• Identifying cancer-promoting gene modules in large datasets is a significant challenge.

Purpose of the Study:

• To develop an accurate and effective network-based method for detecting cancer-associated dysregulated gene modules.
• To improve the identification of key molecular players in cancer progression.

Main Methods:

• Construction of an undirected-weighted gene network incorporating gene exclusivity, coverage, and topology.
• Development of a comprehensive evaluation function for optimal cluster number selection.
• Application of K-means clustering to identify dysregulated modules.

Main Results:

• Netkmeans demonstrated superior statistical significance and biological relevance compared to IBA and CCEN methods.
• Netkmeans achieved higher accuracy, precision, and F-measure than MCODE, CFinder, and ClusterONE.
• Identified modules are critical for cancer generation, development, and progression.

Conclusions:

• Netkmeans is a powerful tool for identifying functionally significant dysregulated gene modules in cancer.
• These findings support Netkmeans's role in precise cancer diagnosis, treatment, and drug development.