Nitro-aromatic compounds molecular descriptors

Helguera-Morales et al. [83] used a topological sub-structural molecular design (TOPS-MODE) approach to predict the carcinogenicity of 48 nitro-aromatic compounds in female rodents. Topological descriptors are derived mainly from knowledge of the connectivity between atoms within a molecule, and are based to some extent on information on atom types and their electronic environment. The model was able to describe 79.1% of the experimental data. It was found that the carcinogenic activity of the compounds analyzed increases in the presence of a primary... 

Fig. 4. A plot of the two largest principal components of the 312 compounds and the 871 molecular descriptors comprising the training set. 1 = macrocyclic nonmusk, 2 = aromatic nitro nonmusks, 3 = macrocyclic musks, and 4 = aromatic nitro musks. The plane defined by the two largest principal components accounts for 39% of the total cumulative variance.

---