Reconstructing targetable pathways in lung cancer by integrating diverse omics data

Area of Science:

• Oncology
• Molecular Biology
• Bioinformatics

Background:

• Global multi-omics profiling offers insights into cancer signaling networks.
• Understanding oncogene-associated networks is crucial for targeted therapies.
• Non-small cell lung cancer (NSCLC) often involves KRAS mutations, driving dependency.

Purpose of the Study:

• To reconstruct targetable signaling networks associated with KRAS dependency in NSCLC.
• To integrate multi-omics data (transcriptome, proteome, phosphoproteome) for network analysis.
• To identify novel therapeutic targets in KRAS-dependent NSCLC.

Main Methods:

• Developed a two-step bioinformatics strategy to integrate multi-omics data.
• Defined an 'abundance-score' combining transcript, protein, and phospho-protein levels.
• Utilized the Prize Collecting Steiner Tree algorithm to identify functional sub-networks.

Main Results:

• Identified three key modules centered on KRAS/MET, LCK/PAK1, and β-Catenin.
• Validated protein activation within these modules in KRAS-dependent cells.
• Demonstrated LCK's critical role in proliferation of KRAS-dependent NSCLC, but not KRAS-independent NSCLC.

Conclusions:

• LCK is a critical gene for proliferation in KRAS-dependent NSCLC.
• LCK represents a potential druggable target for KRAS-dependent lung cancers.
• Multi-omics data integration provides a powerful approach for identifying cancer-specific targets.