A data driven expert system for toxicological read across

Development of new drug currently takes 10 - 12 years with costs of around 2 billion Euro. Main reasons for failures comprise lack of efficacy and unforseen toxicity. For the latter, costs are estimated at 250 mio Euro/compound. Current approaches to minimise the of failures due to toxicity comprise the so called toxicological read across. Briefly, toxicologists query the avilable literature for compounds which are structurally similar to their development candidate in order to retrieve information on potential hazards. In our large scale collaborative industry-academia EU-projecs, which aim at the development of computational approaches for toxicity prediction, we realised that this is done mostly by manual searches. This is time consuming and prone to errors.

With ToxPHACTS we integrate our expertise in computational toxicology and semantic data integration to offer an in silico expert system which will help pharmaceutical companies to foresee possible toxologies of new development candidates. In contrast to the current process for toxicological read across, ToxPHACTS

  • uses innovative ways of similarity searching, such as artificial intelligence driven bioisosteric similarity
  • allows complex queries across multiple, semantically integrated data sources such as the Open PHACTS Discovery Platform
  • provides advanced visualisation tools for rapid and easy analysis of read across search results

ToxPHACTS thus combines highly innovative similarity searching with the power of semantically integrated life science data. This brings toxicological read across to the desktop of every toxicologist, allowing them to focus on the analysis of the results rather than wasting a lot of time for searching. With its cutting edge search and analysis tools, ToxPHACTS will help to save time, to save money, and to reduce animal experiments.

The ToxPHACTS team is headed by Prof. Dr. Gerhard Ecker, professor of Pharmacoinformatics at the University of Vienna. Further members are Jana Gurinova, data scientist, Bernhard Knasmüller, software engineer, Lars Richter, chemoinformatician, Melanie Grandits, specialist for cluster architecture, and Anika Dangl, our project manager. They are currently all members of the Pharmacoinformatics Research Group, University of Vienna. Financial issues will be covered by Florian Ecker, tax advisor, and marketing and sales is done by Johanna Lawrence, Johanna Lawrence Design and Marketing, CO. The team has already performed a series of customer interviews which form the basis for a first lash-up of ToxPHACTS. A first functional prototype of ToxPHACTS is supposed to be released by mid 2018. ToxPHACTS will be developed by Phenaris, a spin-off company of the University of Vienna in the stage of being founded.