Interactive predictive toxicology with Bioclipse and OpenTox - Open Source Software in Life Science Research - page 37

Various data types are used by the various predictive
models described in Table 2.2 to provide detailed
information about what aspects of the molecules
contributed to the decision on the toxicity

Table 2.3

Structural alerts A substructure that has been associated with an

alert, in our case a chemical liability. Can be
implemented in many ways, most common is using
SMARTS patterns. Also referred to as toxicophore if
alerting for toxicity.

Signature alerts A type of structural alert implemented using

Signatures [26].

Signature signifi cance A QSAR model which is capable of, apart from a

prediction, to return the most signifi cant signature in
the prediction [27].

Exact match An identical chemical structure was found in the

training data. Can be implemented by InChI and
Molecular Signature.

Near Neighbor match A similar chemical structure was found in the

training data. Commonly implemented by binary
fi ngerprint and within a certain Tanimoto distance.

2.3 Use Case 1: Removing toxicity
without interfering with pharmacology

From the TCAMS data set, we select the compound TCMDC-135308 for
further investigation. TCMDC-135308 is similar (Tanimoto=0.915) to
quinazoline 3d (both shown in Table 2.4), a potent human-TGF-β1
inhibitor (Transforming Growth Factor-β1). TCMDC-135308 inhibits
growth of the P. falciparum strain 3D7 by 98% at a concentration of
2 μM (XC50=700 nM). It is also active against the multidrug-resistant P.
falciparum strain DD2, inhibiting its growth by 63% at a concentration
of 2 μM. The compound has not shown signifi cant growth inhibition of
human HepG2 cells (5% at a concentration of 10 μM).
We import TCMDC-135308 using its SMILES to Bioclipse, log in to
OpenTox and move to the Decision Support Perspective, where we run
all models. It turns out TCMDC-135308 contains a CPDB Signature
Alert for Carcinogenicity (see Figure 2.4). Signature Alerts in Bioclipse
Decision Support are discriminative signatures identifi ed according to

￿ ￿ ￿ ￿ ￿