PRIORI-T

PRIORI-T prioritizes candidate disease genes for rare diseases by extracting disease–phenotype–gene associations from MEDLINE abstracts and applying the GCAS (Gene-Centric Association Scoring) algorithm.

Key Features:

• Data Source Utilization: Extracts rare disease correlation pairs involving diseases, phenotypes, and genes directly from MEDLINE abstracts.
• Information Propagation Algorithm: Constructs an association network using the GCAS (Gene-Centric Association Scoring) algorithm to propagate associations for gene prioritization.
• Phenotype-Driven Prioritization: Uses phenotype-driven approaches to prioritize genes based on extracted disease–phenotype associations.
• Performance Evaluation: Evaluated on phenotype descriptions from 230 real-world rare disease clinical cases sourced from recent publications.
• Comparative Performance: Demonstrated superior gene prioritization compared to DisGeNET, Open Targets, HANRD, and Orphamizer.
• Quality of Associations: Contains qualitatively better associations than those found in DisGeNET and Open Targets.
• Causal Gene Identification: Identified causal genes within the Top-50 for over 40% of cases and within the Top-300 for more than 72% of cases.

Scientific Applications:

• Rare Disease Research: Supports studies of rare diseases by prioritizing candidate genes when curated resources are sparse or incomplete.
• Gene Prioritization: Provides high-quality disease–gene associations to aid candidate gene ranking for downstream experimental validation.

Methodology:

Extracts associations from MEDLINE abstracts, builds an association network using the GCAS algorithm, and applies phenotype-driven gene prioritization.