Typing tumors using pathways selected by somatic evolution.

Area of Science:

• Genomics
• Cancer Biology
• Bioinformatics

Background:

• Current cancer genome analysis relies on broad pathway maps, which include irrelevant interactions.
• This noise obscures functionally significant cancer pathways and driver mutations.

Purpose of the Study:

• To develop a method for creating tumor-specific molecular networks.
• To improve the identification and classification of cancer subtypes based on evolutionary signatures.

Main Methods:

• Constructed stringent, tumor-specific protein interaction networks using signatures of epistatic selection.
• Applied the evolutionarily selected pathway (ESP) framework to analyze major cancer genome atlases.
• Derived a hierarchical classification of tumor subtypes linked to characteristic mutated pathways.

Main Results:

• Identified clinically prognostic and predictive pathways, such as TP53-AXIN-ARHGEF17 in liver cancer and CYLC2-STK11-STK11IP in lung cancer.
• Validated these findings in independent patient cohorts.
• Demonstrated that the ESP framework significantly refines the definition of cancer pathways and subtypes.

Conclusions:

• The ESP map offers a more accurate and robust approach to analyzing cancer genomes.
• This framework improves the biological and clinical relevance of identified cancer pathways and subtypes.
• Enhances the potential for precision oncology by better defining tumor characteristics.