Identification of key processes underlying cancer phenotypes using biologic pathway analysis

Area of Science:

• Genomics
• Molecular Biology
• Cancer Research

Background:

• Cancer is a group of gene-based diseases caused by disruptions in fundamental biological processes.
• Current methods for linking disease phenotypes to molecular network alterations are indirect and unsystematic.
• A systematic approach is needed to map biological pathways to clinical features in cancer.

Purpose of the Study:

• To objectively identify biological pathways associated with malignancy, staging, and outcome in cancer.
• To systematically evaluate differences in pathway activity and interaction consistency.
• To develop a method for mapping genome-wide data to clinical features.

Main Methods:

• Analysis of large collections of publicly accessible genome-wide gene expression data.
• Systematic evaluation of differences in activity and consistency of interactions within canonical biological processes.
• Application of an analytic approach to identify pathways distinguishing tumor types, grades, and phenotypes.

Main Results:

• Identified small, common sets of pathways (Trka Receptor, Apoptosis response to DNA Damage, Ceramide, Telomerase, CD40L, Calcineurin) that robustly distinguish tumor types from normal samples.
• These pathways accurately predict tumor grade and phenotypes like estrogen receptor status and p53 mutation state.
• Identified pathways demonstrated equal or superior performance compared to existing phenotypes in predicting cancer outcomes.

Conclusions:

• The identified pathways provide robust biomarkers for cancer classification and prognostication.
• This approach enables the use of genome-wide data to link key biological processes to critical clinical features in cancer.
• The findings offer a systematic means to understand the molecular underpinnings of cancer heterogeneity and progression.