Capturing and curating in vitro data from diverse sources to support MIE investigation for hepatotoxicity
Many models for the prediction of hepatotoxicity can be viewed as ‘top-down’, based on extrapolations from in vivo data. For modelling purposes, the utility of in vivo hepatotoxicity data is limited for a number of reasons. These include the paucity of available data for large numbers of compounds, the use of heterogeneous assays and their resulting responses, and the contribution of ADME factors. With such data it is also difficult to disentangle the various molecular initiating events (MIEs) and subsequent pathways which may contribute to the eventual outcome. We reviewed the published literature on MIEs and adverse outcome pathways (AOPs) related to hepatotoxicity, to try to identify which existing in vitro assay data could be mapped to the available MIEs. Published sources were then investigated for collections of in vitro data of interest. Particularly relevant assays include those that test for mitochondrial liabilities and for transporter interaction (BSEP inhibition, MRP2). We collated and curated this data to create a dataset of 8390 compounds. The major single data source for this information was found to be the Tox21 project providing over 7000 compounds. We present a review of published MIEs for hepatotoxicity, and our own work in creating a single database of in vitro results that can support the investigation of MIEs for hepatotoxicity prediction.
Poster presented by Mukesh Patel at FutureTox III, Arlington, USA; 20th November 2015.