A new version of the eTRANSAFE preclinical text mining system, PretoxTM 2.0, was presented in a virtual workshop on the 22nd of November 2021 by members of the project Work Package 5.
The workshop was organized as a brainstorming session to obtain end-user feedback from toxicology experts. The insights gathered during the event will be used to feed the continuous improvements of the preclinical text mining system and will be incorporated in subsequent versions.
Javier Corvi (Barcelona Supercomputing Centre – BSC) introduced the system and its functionalities to the audience:
“Text mining technologies provide a solution to analyse the large and growing number of unstructured text documents and extract relevant information from them. In case of toxicology reports, the abnormal observations or findings recorded on the test subjects after the administration of a compound during a given study are the kind of relevant information we wish to retrieve.
The PretoxTM system has two main modules; the text mining pipeline that detects and retrieves treatment-related findings from toxicological reports using NLP (Natural Language Processing) techniques; and the PretoxTM Web Application that presents the extracted findings to the experts for validation and curation.
In other words, the main goal of the PretoxTM system is to assist and help the toxicology experts to detect treatment-related findings in the toxicology reports and present this information in a user-friendly interface via the SR domain.”
We thank all workshop participants for their valuable insights and look forward to seeing the fruits of these interactions in further improvements of the preclinical text mining system!