Target safety assessments help understand the role of a drug target in normal physiology, and any potential unintended adverse consequences (safety liabilities) of target modulation (Brennan 2017).
We perform manual curation of experimental data available from the literature and other resources of toxicity. Target safety data is subdivided into three categories:
- Known side effects
- Safety risk information
- Non-clinical experimental toxicity
Known side effects were manually curated from Bowes et al. (2012), Urban et al. (2012), Lamore et al. (2017), Lynch et al. (2017), and HeCaToS.
Gene names were converted into Ensembl Gene IDs, main organs (or systems affected) were mapped to the Uber-anatomy ontology, and source terms (e.g. atrial fibrillation, cardiomyopathy) were mapped to the Experimental Factor Ontology.
Safety risk information was manually curated from two sources, namely Force and Kolaja (2011) and HeCaToS.
Gene names were converted into Ensembl Gene IDs and main organs (or systems affected) were mapped to the Uber-anatomy ontology.
Raw data was retrieved from Exploring ToxCast Data: Downloadable Data on 3rd December 2019. This dataset comprises experimental toxicology data from publications selected by Tox21 as well as industry legacy reports. Tox21 public data is available in the NIH-National Center for Advancing Translational Sciences website. The data curation steps involved selecting target-centric experimental data points and excluding any data that did not report any toxicity parameters. Gene names were mapped to Ensembl identifiers and identity disambiguation was carried out using the UniProt identity mapping tool, the Ensembl website with cross reference with HGNC.
The eTOX project is an Innovative Medicines Initiative dedicated to extracting and sharing preclinical study data (Cases et al. 2014). This dataset comprises toxicology data from scientific publications as well as legacy toxicology data from 13 participant pharmaceutical companies. Experimental information for human data targets from the E-TOX_Protein_compilation dataset was retrieved on 25th November 2019. The data was standardised for identifiers, where all targets were mapped to Ensembl gene IDs. Experimental information with toxicity related tissue and organ system was standardised according to the CDISC SEND Controlled Terminology.
Data on target safety is available for 439 targets and you can access this data via the:
- Open Targets Platform website e.g. in the profile pages for PAX6 and TBXA2R
- Open Targets Platform REST API, using the
private/targetendpoint - Data Download page
{% hint style="info" %} Looking for significant adverse drug reactions (ADRs) instead? Check our pharmacovigilance page. {% endhint %}