PHAEDRA

PHAEDRA is a text mining tool that supports curators of pharmacovigilance databases by automatically identifying relevant evidence on drug effects and their interactions from textual data. It provides a semantically annotated corpus of 597 MEDLINE abstracts, which encodes rich information on drug effects and their interactions, and has high levels of inter-annotator agreement. TM tools can be trained based on its rich labels to automatically recognize drug effects in text. The corpus and annotation guidelines are available for use.

Topic

Cheminformatics;Drug discovery;Medicinal chemistry;Natural language processing

Detail

  • Operation: Text mining

  • Software interface: Web user interface

  • Language: -

  • License: Attribution-ShareAlike 4.0 International

  • Cost: Free

  • Version name: -

  • Credit: EPSRC, MRC (MMPathIC project), the BBSRC (EMPATHY project), the EC-H2020 (OpenMinTeD project).

  • Input: -

  • Output: -

  • Contact: paul.thompson@manchester.ac.uk

  • Collection: -

  • Maturity: -

Publications

  • Annotation and detection of drug effects in text for pharmacovigilance.
  • Thompson P, et al. Annotation and detection of drug effects in text for pharmacovigilance. Annotation and detection of drug effects in text for pharmacovigilance. 2018; 10:37. doi: 10.1186/s13321-018-0290-y
  • https://doi.org/10.1186/s13321-018-0290-y
  • PMID: 30105604
  • PMC: PMC6089860

Download and documentation

< Back to DB search