The JRC dataset (doi: 10.1016/j.tiv.2016.07.014) includes reported outcomes in chemico, in vitro and in vivo tests for skin sensitisation. The Tox21 dataset was downloaded from the EPA CompTox Chemicals Dashboard (available at: https://comptox.epa.gov/dashboard/chemical_lists/TOX21SL) and was filtered to retain substances that were discrete organics with normalized structures, i.e. only substances that had an associated QSAR-ready SMILES record.
This dataset is associated with the following publication:
Nelms, M., R. Lougee, D. Roberts, A. Richard, and G. Patlewicz. Comparing and contrasting the coverage of publicly available structural alerts for protein binding. Computational Toxicology. Elsevier B.V., Amsterdam, NETHERLANDS, 12: 100100, (2019).