In-silico interaction-resolution pathway activity quantification and application to identifying cancer subtypes.

Area of Science:

• Computational biology
• Genomics
• Cancer research

Background:

• Identifying cancer subtypes is crucial for developing personalized therapeutics due to varying disease origins.
• Current subtype identification often relies on gene-specific signatures, overlooking complex gene interactions.
• Network-driven biological functions and differential gene interactions in molecular contexts are gaining research interest.

Purpose of the Study:

• To propose an in-silico method for quantifying pathway activities at the genetic interaction level.
• To develop a method for calculating sample discrepancies based on quantified pathway activities.
• To identify novel cancer subtypes and generate hypotheses regarding their underlying mechanisms.

Main Methods:

• Developed an in-silico approach to quantify pathway activities, incorporating genetic interaction resolution.
• Created a discrepancy measure to compare samples based on quantified pathway activities.
• Applied the method to cluster melanoma gene expression data.

Main Results:

• Identified two potential melanoma subtypes characterized by distinct pathway activities.
• Observed significantly different survival patterns between the identified patient groups.
• Investigated specific pathways, suggesting potential mechanisms driving the identified subtypes.

Conclusions:

• The proposed method effectively models pathway activities with genetic interaction resolution.
• Novel potential disease subtypes were proposed, offering insights into subtype-specific genetic interactions.
• The findings provide a foundation for generating hypotheses on disease mechanisms and developing targeted interventions.