POKR: Building a Computable Heterogeneous Knowledge Resource for Precision Oncology.

Area of Science:

• Oncology
• Bioinformatics
• Genomics

Background:

• Precision oncology aims to improve cancer patient outcomes through targeted therapies.
• Existing knowledge bases like CIViC and OncoKB provide valuable genetic-treatment evidence but are analyzed separately.
• An integrated approach is needed to leverage the full potential of these resources for drug discovery.

Purpose of the Study:

• To develop a heterogeneous precision oncology knowledge resource (POKR) by integrating CIViC and OncoKB.
• To enable computable and measurable associations among genes, drugs, and diseases.
• To support timely drug repurposing discovery for life-threatening cancers.

Main Methods:

• Integrated data from CIViC and OncoKB into a unified knowledge resource named POKR.
• Developed graph embeddings for POKR to represent biomedical entities and their relationships.
• Trained POKR graph embeddings to enable computational analysis of associations.

Main Results:

• Successfully built a heterogeneous precision oncology knowledge resource (POKR).
• Established computable associations among biomedical entities within POKR.
• Demonstrated the potential for enhanced drug repurposing discovery.

Conclusions:

• The integrated POKR resource offers a novel approach to precision oncology knowledge management.
• POKR facilitates the discovery of new therapeutic strategies by connecting genetic information with drug treatments.
• This pilot study provides a foundation for further development and application of integrated knowledge bases in cancer research.