Pharos

Pharos aggregates and organizes heterogeneous biomedical datasets from the NIH Illuminating the Druggable Genome (IDG) initiative and the Target Central Resource Database (TCRD) to enable identification and prioritization of understudied protein targets that can be modulated by small molecules or biologics.

Key Features:

• Data integration with TCRD: Integrates the Target Central Resource Database (TCRD) as a central data source for target-centric aggregation.
• Expanded data types: Incorporates over 25 additional data types beyond core TCRD content.
• Protein–protein interactions: Includes human and viral-human protein-protein interaction datasets.
• Protein–disease and phenotype associations: Aggregates associations between proteins and diseases or phenotypes.
• Drug-induced gene signatures: Integrates drug-induced gene signature data linked to protein targets.
• Machine learning-ready outputs: Produces datasets formatted for machine learning workflows.
• Target ranking by data volume: Enables ranking and prioritization of targets based on the amount of supporting data.

Scientific Applications:

• Understudied protein illumination: Supports identification and study of poorly characterized proteins within the druggable genome.
• Target prioritization for drug discovery: Facilitates prioritization of protein targets for small-molecule or biologic therapeutic development.
• Computational and ML analyses: Provides machine learning-ready datasets for computational target-ranking and integrative analyses.

Methodology:

Integrates data from TCRD and over 25 additional data types—including human and viral-human protein–protein interactions, protein–disease/phenotype associations, and drug-induced gene signatures—and outputs machine learning-ready datasets to support target ranking by data volume.