Comparative Analysis of Public Knowledge Bases for Precision Oncology

Area of Science:

• Genomics
• Bioinformatics
• Precision Medicine

Background:

• Precision oncology relies on accurate genetic variant and therapeutic association data.
• Knowledge bases (KBs) compile this information from scientific literature.
• Several KBs exist, necessitating a comparison of their utility.

Purpose of the Study:

• To quantitatively and qualitatively compare seven major cancer knowledge bases.
• To assess the comprehensiveness and unique contributions of each KB.
• To evaluate the impact of KB usage on clinical decision-making in precision oncology.

Main Methods:

• Downloaded and standardized data from seven KBs into a common format.
• Normalized gene and drug identifiers for consistent analysis.
• Analyzed gene-drug associations using data from 113 cancer patients presented at a Molecular Tumor Board (MTB).

Main Results:

• KBs show substantial overlap but also contain unique, valuable information.
• Individual KBs vary in their coverage of genes and gene-drug associations.
• Using multiple KBs significantly improved the retrieval of clinically relevant information compared to single sources.

Conclusions:

• No single KB currently encompasses all relevant information for precision oncology.
• Each KB possesses distinct strengths and weaknesses.
• Integrating data from multiple KBs is essential for comprehensive and accurate clinical decision-making.