Structure-activity relationship oriented languages for chemical structure representation

Computer aided prediction of biological activity spectra: Evaluating versus known and predicting of new activities for thiazole derivatives

“…To evaluate the applicability of the approach to the thiazole derivatives, the results of prediction are compared to experimental data for some compounds from the set, studied as 3,4,5,6,7,8,9,10,11,12,14,15,17,18,19,20,26,27,28,29 non-steroid antiinflammatory (78 compounds); local anaesthetics (6 compounds); antioxidants (5 compounds) tests. These data are given in Table II. In the experiment the compound is considered as being active if its potency is no less than the potency of the appropriate reference drug (the experimental errors are taken into account).…”

“…Thus, in the simultaneous prediction of the wide range of the activity kinds for non-congeneric compounds it is necessary to use the description, which would be exactly sufficient to achieve a consistent prediction result but would not be too sensitive to ignore some random regularities. In earlier versions of PASS, the Substructure Superposition Fragment Notation (SSFN) [28] was used. In our paper published recently [22], we described the substructure descriptors called "Multilevel Neighborhoods of Atoms" (MNA).…”

Probabilistic Approaches in Activity Prediction

“…3,67,136-139 Earlier 29 we applied the Substructure Superposition Fragment Notation (SSFN) codes. 140 But SSFN, like many other structural descriptors, reflects rather abstraction of chemical structure by the human mind than the nature of the biological activity revealed by chemicals. The Multilevel Neighborhoods of Atoms (MNA) descriptors 91,141,142 have certain advantages over SSFN.…”

Computer-assisted examination of compounds for common three-dimensional substructures

“…Consider the three-atom nodeset 9, 10, N o d e s e t W a r b u r g a n a l 7 9 16 for warburganal (1) shown in Figure 8. For these three atoms, there are three connections to be described.…”

“…Recently the Stanford group has developed an array of sophisticated method^.^ Several other groups are also very active in advancing this important area. [8][9][10][11][12][13][14][15] In the second approach, there exist empirical rules such as those formulated by Grant boxylic acids22, along with numerous others observed for other functional g r o~p s .~~,~~ Clerc and Pretsch have devised general additive rules for 28 functional groups.25 Dubois has used a topological parameter to model the alkyl environment.z6 Levy and Nelsonz7 and E j~h a r t~~, *~ have proposed substitution methods whereby the 13C shifts are first estimated for the hydrocarbons, and heteroatoms are substituted later. Although these rules have varying accuracy, they serve as good starting points for spectral interpretation, especially when simple analogues cannot be located in the spectral libraries.…”