Literature evidence in open targets - a target validation platform

Area of Science:

• Biomedical Informatics
• Computational Biology
• Drug Discovery

Background:

• The Open Targets platform integrates diverse evidence for drug target identification and validation.
• The Europe PMC literature component specifically extracts target-disease associations from scientific documents.
• This component utilizes expert-defined rules and heuristic information to rank document relevance for target validation.

Purpose of the Study:

• To present the Europe PMC literature component of the Open Targets platform.
• To detail the methodology for identifying and ranking target-disease associations from scientific literature.
• To highlight the value of literature-derived evidence in drug target validation.

Main Methods:

• Text mining of over 26 million PubMed abstracts and 1.2 million Open Access full-text articles.
• Rule-based system utilizing expert-provided heuristic information for confidence scoring.
• Ranking of documents based on their credibility and relevance to specific target-disease associations.

Main Results:

• Identification of 1,168,365 distinct target-disease associations.
• Over 850,000 associations were exclusively identified through literature mining, underscoring its unique contribution.
• The component has continuously provided updated data since December 2015.

Conclusions:

• The literature component delivers highly relevant scientific literature for specific targets and diseases.
• It enhances the Open Targets platform by incorporating crucial text-mining evidence.
• Users can explore this data visually or download it for further analysis.