Elucidating synergistic dependencies in lung adenocarcinoma by proteome-wide signaling-network analysis

Area of Science:

• Biochemistry
• Systems Biology
• Oncology

Background:

• Understanding drug interactions requires mapping signaling pathways, often based on limited literature data.
• Reconstructing signaling networks from large-scale molecular data is challenging compared to other biological networks.

Purpose of the Study:

• To develop a novel algorithm for systematic inference of protein kinase pathways.
• To apply this algorithm to lung adenocarcinoma (LUAD) phosphotyrosine data for network reconstruction.
• To validate the model's accuracy and utility in predicting drug response.

Main Methods:

• Developed a novel algorithm for protein kinase pathway inference.
• Applied the algorithm to mass spectrometry-based phosphotyrosine profiling data from 250 LUAD samples.
• Validated inferred network components using SILAC assays.

Main Results:

• Inferred a LUAD-specific network of 43 tyrosine kinases (TKs) and 415 substrates.
• Achieved >60% validation accuracy, identifying novel substrates for EGFR and c-MET TKs.
• Demonstrated the model's ability to predict drug response, including synergistic inhibitor activity.

Conclusions:

• The data-driven approach enables systematic reconstruction of signaling pathways.
• The inferred network aids in understanding LUAD-specific signaling and identifying therapeutic targets.
• This model supports precision oncology by predicting individual-level drug responses.