Interpretation of QSAR Models Based on Random Forest Methods.

A new algorithm for the interpretation of Random Forest models has been developed. It allows to calculate the contribution of each descriptor to the calculated property value. In case of the simplex representation of a molecular structure, contributions of individual atoms can be calculated, and thus it becomes possible to estimate the influence of separate molecular fragments on the investigated property. Such information can be used for the design of new compounds with a predefined property value. The proposed measure of descriptor contributions is not an alternative to the importance of Breiman's variable, but it characterizes the contribution of a particular explanatory variable to the calculated response value.

Synthesis and biological activity of 7H-benzo[4,5]indolo[2,3-b]-quinoxaline derivatives.

New 7-(2-aminoethyl)-7H-benzo[4,5]indolo[2,3-b]quinoxalines (13-20) were synthesized with high yields starting from 3H-benzo[e]indole-1,2-dione. These compounds were screened for the cytotoxicity, anti-viral activity, interferon inducing ability and DNA affinity compared with the corresponding 6-(2-aminoethyl)-6H-indolo[2,3-b]quinoxaline derivatives (1-12). It was shown, that compounds 13-20 bind to DNA stronger (lg Кa=6.23-6.87) than compounds 1-12 (lg Кa=5.57-5.89). Anti-viral activity is significantly reduced with annulations of benzene ring in Indoloquinoxaline moiety 13-20.

Synthesis, cytotoxicity, antiviral activity and interferon inducing ability of 6-(2-aminoethyl)-6H-indolo[2,3-b]quinoxalines.

New 6-(2-aminoethyl)-6H-indolo[2,3-b]quinoxalines were synthesized with high yields using bromoethylisatin and 6-(2-bromoethyl)-6H-indolo[2,3-b]quinoxaline as intermediates. These compounds were screened for the cytotoxicity, antiviral activity and interferon inducing ability. It was shown, that tested 6-(2-aminoethyl)-6H-indolo[2,3-b]quinoxalines are low toxic potent interferon inducers and antivirals. Morpholine and 4-methyl-piperidine derivatives appeared as the most active antivirals and the least cytotoxic in the investigated series.

Hierarchic system of QSAR models (1D-4D) on the base of simplex representation of molecular structure.

In this work, a hierarchic system of QSAR models from 1D to 4D is considered on the basis of the simplex representation of molecular structure (SiRMS). The essence of this system is that the QSAR problem is solved sequentially in a series of the improved models of the description of molecular structure. Thus, at each subsequent stage of a hierarchic system, the QSAR problem is not solved ab ovo, but rather the information obtained from the previous step is used. Actually, we deal with a system of solutions defined more exactly. In the SiRMS approach, a molecule is represented as a system of different simplex descriptors (tetratomic fragments with fixed composition, structure, chirality and symmetry). The level of simplex-descriptor detail increases consecutively from 1D to 4D representations of molecular structure. It enables us to determine the fragments of structure that promote or interfere with the given biological activity easily. Molecular design of compounds with a given level of activity is possible on the basis of SiRMS. The efficiency of the method is demonstrated for the example of the analysis of substituted piperazines affinity for the 5-HT1A receptor.