iGPSe: a visual analytic system for integrative genomic based cancer patient stratification

Area of Science:

• Genomics
• Bioinformatics
• Translational Informatics

Background:

• Cancer is a heterogeneous disease with diverse subtypes.
• Subtypes exhibit varying genetic profiles, phenotypes, and clinical outcomes.
• Integrative genomics (panomics) aims to identify biomarkers for patient stratification.

Purpose of the Study:

• To present the Interactive Genomics Patient Stratification explorer (iGPSe) system.
• To reduce the computational burden of exploring complex integrative genomics data.
• To facilitate the identification of integrative biomarkers for patient stratification.

Main Methods:

• Developed a visual analytic system (iGPSe).
• Integrated unsupervised clustering with graph and parallel sets visualization.
• Incorporated survival analysis for direct comparison of clinical outcomes.

Main Results:

• iGPSe effectively reduces the computing burden for researchers.
• Explored combinations of gene expression (mRNA) and microRNA features.
• Identified potential combined markers for survival prediction using TCGA breast cancer data.

Conclusions:

• Visualization is crucial for patient stratification.
• Visual tools enable feature selection across diverse datasets.
• This work highlights the importance of visualization in translational informatics.