Identification of Pharmacologically Tractable Protein Complexes in Cancer Using the R-Based Network Clustering and...

Area of Science:

• Computational biology
• Network analysis
• Cancer research

Background:

• Multiomics assays identify biological network entities.
• Molecular Complex Detection (MCODE) is a key algorithm for network subclusters.
• Existing MCODE implementation on Cytoscape has limitations for large-scale, simultaneous analyses.

Purpose of the Study:

• To implement the MCODE algorithm in R, creating a package named MCODER.
• To enable simultaneous analysis of multiple, large-scale omics datasets within the R environment.
• To apply MCODER for identifying therapeutically relevant protein-protein interactions in cancer subtypes.

Main Methods:

• Developed the MCODER package by implementing the MCODE algorithm in R.
• Utilized MCODER for analyzing multiple omics data sets concurrently.
• Applied MCODER to detect protein-protein interactions in molecular subtypes of ovarian and colorectal tumors.

Main Results:

• MCODER facilitates efficient analysis of multiple omics datasets.
• Identified elevated collagen-integrin protein complex production in an epithelial-mesenchymal transition subtype common to both ovarian and colorectal cancers.
• This finding highlights a potential therapeutic vulnerability in this specific cancer subtype.

Conclusions:

• MCODER provides a flexible and efficient R-based solution for network biology analyses.
• The collagen-integrin complex is a promising target for pharmacological intervention in specific molecular subtypes of ovarian and colorectal tumors.
• This study demonstrates the utility of MCODER in uncovering biologically significant molecular interactions in cancer.