The COMPACT predictions have also been compared with published Ames mutagenicity data and with our own Hazardexpert predictions for carcinogenicity.
To validate the developed models two external data sets, AID1189 and AID1194, with mutagenicity data were tested showing 62% accuracy with 67% precision and 65% ROC area and 91% accuracy, 91% precision with 96.3% ROC area respectively.
A set of chemicals specified with mutagenicity data was retrieved, and necessary information was extracted and transferred to the working file.
Previously the Benchmark data set was the largest mutagenicity data set containing more than 6000 molecules classified as mutagens and non-mutagens.
This work included 3 training data sets: Set 1 (Bursi mutagenicity data set) having a total of 4337 compounds, Set 2 (Benchmark data set) with 6512 compounds and Set 3 which was a combination of Set 1 and Set 2 containing 8208 compounds after removing the duplicate structures based on the canonical smiles of the Set 1 and Set 2 using Pipeline Pilot [21].
In pooling data sets on sucrose, invert sugar, honey, and HFCS from various studies and laboratories (Roemer et al., 2010; Coggins et al., 2011), mutagenicity data for these two testing conditions did not show any increase in activity with increasing single sugar application levels (TPM-based comparison, Figure 5A and B).
Mutagenicity data from these experiments support the theory of formation of direct-acting mutagens, probably due to nitration.
Kazius et al. provided an analysis of mutagenicity data correlating chemical substructures to mutagenicity [21, 27, 28].
Table 1 Aryl-amine mutagenicity data sets considered in this study.
In addition, toxic compounds were filtered out using a categorical SVM model for prediction of mutagenicity trained on Ames-test mutagenicity data (5-fold ROC score: 0.8948).
