| Debnath, A.K. Lopez de Compadre, R.L., Debnath, G.,
Shusterman, A.J., and Hansch, C. (1991).
Structure-activity relationship of mutagenic
aromatic and heteroaromatic nitro compounds.
Correlation with molecular orbital energies
and hydrophobicity.
J. Med. Chem. 34:786-797.
R. King and S. Muggleton and A. Srinivasan and M. Sternberg
Structure-activity relationships derived by machine
learning: the use of atoms and their bond connectives
to predict mutagenicity by inductive logic programming.
Proceedings of the National Academy of Sciences,
vol.93:438-442, 1996.
A. Srinivasan and S. Muggleton and R. King and M. Sternberg
Theories for mutagenicity: a study of first-order and
feature based induction. (1996), J.Artificial Intelligence,
vol. 85 (1,2):277-299.
|
| The prediction of mutagenesis is important as it is relevant to understanding and prediction of carcinogenesis. Not all compounds can be empirically tested for mutagenesis, e.g. antibiotics. The considered data have been collected with the intention to search for a method how to predict the mutagenicity of a set of 230 aromatic and heteroaromatic nitro compounds, the dataset is based on the results published in [Debnath et al (1991)].
Mutagenicity is measured by the Ames test using S. typhimurium TA98. Of the 230 compounds 138 with positive levels of log mutagenicity are labeled as active and they constitute positive examples. The remaining 92 compounds represent the negative examples. Four attributes are provided for the analysis of the components, namely
- hydrophobicity,
- energy level of the lowest unoccupied molecular orbit,
- two boolean valued attributes identifying components with "three or more benzol rings" and compounds termed accenthryls.
The considered nitro compounds are more heterogeneous structurally than any of those in the other ILP datasets concerning chemical structure activity. Results of relevance to ML community are available in [Srinivasan 94, 95a, 95b], relevant chemical results are in [King 95].
|
and GMD 