Toxicity Predictions - Random Forest Ensemble

Service is currently down to license upgrades

This page uses a 10-model ensemble of random forests to predict whether a structure (specified in SMILES format) is toxic or not, based on 1052 bit BCI fingerprints. The models are built on a subset of the data reported in PubChem assay 463 (Cell Proliferation & Viability [Cytotoxicity]).

Note, this page is a proof-of-concept of model deployment using R web services coupled to a remote R engine.

Paste SMILES, one to a line

Some example SMILES:

    CCOc1ccc(cc1)C2CC(c3cccc(Cl)c3)n4nc(N)nc4N2
    CCOC1(OCC)N=C(N)C2(C#N)C(c3cccc(OC)c3OC)C12C#N
    Nc1nc2NC(CC(c3ccc(Cl)cc3)n2n1)c4ccc(F)cc4
    COc1ccc(cc1)C2CC(c3ccc(Cl)cc3)n4nc(N)nc4N2
    COc1cccc(c1)C2CC(c3ccc(Cl)cc3Cl)n4nc(N)nc4N2
    COc1ccc(cc1)C2CC(c3ccc(Cl)cc3Cl)n4nc(N)nc4N2
    CC(C)COC(=O)C1C2OC3(CN(C4CCCC4)C(=O)C13)C=C2
    CN1C(=O)N(C)C2=C(C1=O)C(NS(=O)(=O)c3ccccc3)(C(=O)N2)C(F)(F)F
    FC(F)(F)C1(NS(=O)(=O)c2ccccc2)N=C3SCCN3C1=O