SIBAR descriptor

We recently introduced the concept of using similarity values as independent variables in QSAR equations. Within this similarity-based SAR (SIBAR), similarity values between training set compounds and a set of reference compounds are calculated and subsequently used as molecular descriptors. The approach for calculating the SIBAR descriptors is outlined in Figure 7.11. 

Calculation of a set of descriptors for both the training set and the reference set. Calculation of similarity values for each compound of the training set to each compound of the reference set this leads to a given number of similarity values (equal to the number of reference compounds used) for each compound of the training set, which are assigned as SIBAR descriptors. 

Bayesian and counter propagation neural networks N = 285 from the literature. 3 classes > lOpM, fold cross-validation, N = 58 newly selected compounds used as a test set as well as the PubChem data N = 193 actives, N = 1703 inactives. Used SIBAR Descriptors. 85-88% classification of hERG blockers. 195... 

Zdrazil, B., Kaiser, D., Kopp, S., Chiba, P., and Ecker, G.E. (2007) Similarity-based descriptors (SIBAR) as tool for QSAR-studies on P-glycoprotein inhibitors influence of the reference set. QSAR ei Combinatorial Science, 26, 669-678. 

---