Drug Normalization for Cancer Therapeutic and Druggable Genome Target Discovery

Area of Science:

• Bioinformatics
• Genomics
• Pharmacology

Background:

• Drug data is represented heterogeneously across knowledge resources, hindering cancer therapeutic target discovery.
• Standardization is needed to integrate diverse datasets for effective drug development.

Purpose of the Study:

• To describe challenges and lessons learned in developing a standards-based drug normalization framework.
• To evaluate this framework for cancer druggable genome datasets.

Main Methods:

• Development of a standards-based drug normalization framework.
• Evaluation of the framework using cancer druggable genome datasets.
• Analysis of data from The Cancer Genome Atlas (TCGA), NCI Thesaurus (NCIt), and PubChem.

Main Results:

• Identified the need for mechanisms to handle spelling errors and irregularities in clinical drug data (e.g., TCGA).
• Demonstrated that NCIt and PubChem annotations serve as crucial normalization layers.
• Showcased the potential of these layers to bridge clinical phenotypes and druggable genome knowledge.

Conclusions:

• A standards-based approach is essential for normalizing heterogeneous drug data.
• Effective normalization facilitates the integration of clinical and genomic data for cancer target discovery.
• NCIt and PubChem provide valuable layers for bridging phenotypes and druggable targets.