Data integration to prioritize drugs using genomics and curated data.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Genomic alterations in drug targets are crucial for personalized cancer treatments.
• Patient selection for targeted therapies relies on molecular alterations.
• Integrating diverse molecular data enhances precision therapy selection but existing methods are limited.

Purpose of the Study:

• To develop a novel data mining algorithm for integrating heterogeneous molecular data.
• To facilitate drug repositioning by stratifying patient samples and prioritizing drug targets.
• To enhance the selection of precision therapies for cancer patients.

Main Methods:

• Constructed a knowledge base integrating public databases, tumor molecular data, signaling pathways, and drug-target information.
• Developed a data mining algorithm to utilize this heterogeneous knowledge base for drug prioritization.
• Applied the framework to analyze 797 breast and ovarian cancer tumors from The Cancer Genome Atlas.

Main Results:

• The algorithm successfully integrated diverse data types for drug prioritization.
• FGFR, CDK, and HER2 inhibitors were prioritized for breast and ovarian cancer datasets.
• Estrogen receptor-positive breast tumors showed potential sensitivity to FGFR inhibitors due to FGFR3 activation.

Conclusions:

• Computational sample stratification effectively identifies potentially sensitive patient populations for targeted therapies.
• The developed framework aids in precision medicine and drug repositioning strategies.
• Source code is publicly available for broader research application.