Background

Prognostic signatures and risk stratification are vital to clinical decision making of treatment options in cancer precision medicine. As patient prognosis remains poor, researchers are seeking to develop improved prognostic signatures using molecular information, such as incorporating long non-coding RNA expression. However, incorporating somatic mutation profiles into prognostic signatures has remained a challenge and is often overlooked due to the sparse and binary nature of somatic mutation data. While existing methods mitigate the sparsity challenge by incorporating pathways and gene-gene network information, efforts to address the binary nature of somatic mutation data is minimal.

Technology Overview

Researchers at the University of Hawaii John A. Burns School of Medicine Bioinformatics Core have developed a continuous somatic mutation profile that leads to robust prognostic signatures for cancers. The novel profiles were evaluated in two cancer types, lung adenocarcinoma and colorectal adenocarcinoma, using controlled whole exome sequencing data from The Cancer Genome Atlas (TCGA). Unlike raw binary or gene-gene network normalized somatic mutation profiles, the continuous somatic mutation profiles led to statistically significant prognostic biomarker signatures in independent validation datasets. Literature, network and pathway analysis of the resulting prognostic biomarker signatures confirm biological relevancy and the potential to discover novel cancer genes.

Benefits

Leads to more statistically robust prognostic biomarker signatures than using raw binary or gene-gene network normalized somatic mutation profiles
Inferences can be made with small sample sizes with continuous somatic mutation profiles
Continuous somatic mutation profiles offer higher sensitivity and larger range of analysis options over binary profiles

Applications

Prognostic tool for specific cancer types
Diagnostic tool for specific cancer types
Identification of somatic mutation biomarkers for drug response

Method for Continuous Somatic Mutation Profiles for Prognostic Biomarker Signatures

IP Status

Seeking