Oncopacket: integration of cancer research data using GA4GH phenopackets.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Data integration is a major hurdle in cancer research.
• Current analyses often necessitate custom software for data preparation.
• Standardized data formats are crucial for advanced analytics.

Purpose of the Study:

• To present a software package for harmonizing genetic and clinical cancer data.
• To implement the GA4GH Phenopacket schema for standardized data representation.
• To facilitate downstream statistical and AI/ML analyses in cancer research.

Main Methods:

• Developed a software package to integrate diverse cancer data types.
• Utilized the GA4GH Phenopacket schema, an ISO standard.
• Integrated demographic, mutation, morphology, diagnosis, intervention, and survival data.
• Applied the software to National Cancer Institute data across 12 cancer types.

Main Results:

• Successfully harmonized genetic and clinical data into the Phenopacket schema.
• Demonstrated the utility of the integrated data for complex analyses.
• Replicated a known association between IDH1 gene mutations and survival in brain cancer patients.

Conclusions:

• The software package effectively addresses data integration challenges in cancer research.
• The Phenopacket schema provides a robust foundation for AI/ML and statistical analyses.
• Standardized data facilitates deeper insights into cancer biology and patient outcomes.