Fatima Adilova and Alisher Ikramov
National University of Uzbekistan, Uzbekistan
Posters & Accepted Abstracts: Med chem
The activity cliff concept is of high relevance for medicinal chemistry. Herein, we explore a concept of â??data set modelabilityâ?, i.e., a priori estimate of the feasibility to obtain externally predictive QSAR models for a data set of bioactive compounds. This concept has emerged from analyzing the effect of so-called â??activity cliffsâ? on the overall performance of QSAR models. Some indexes of â??modelabilityâ? (SALI, ISAC, and MODI) are known already. We extended the version of MODI to data sets of compounds with real activity values. We chose out of 5231 compounds from CHEMBL database, for which activity regarding CA2 protein (Inhibitory activity against human recombinant carbonic anhydrase II) was calculated. The data set divided into some samples, containing 100 and 50 compounds in each. There are 19 real-valued descriptors for each compound in CHEMBL that we used in the calculations. The predictive efficiency of QSAR models is expressed as the correct classification rate by SVM algorithm, which compared with the results of the other two algorithms: algorithm MODI and Voronin algorithm modified by the authors. Comparative analysis of the results performed using Pearsonâ??s correlation coefficient square. Our study showed an extreme lack of evaluation of predictive efficiency of data set only based on â??activity cliffsâ?. In the development of more accurate methods that allow to evaluate the possibility of building of effective models on the data samples, it is necessary to take into account other properties of the sample, and not only the presence (and number) of â??activity cliffsâ?.
Email: fatima_adilova@rambler.ru
Medicinal Chemistry received 6627 citations as per Google Scholar report
Spanish 
Chinese 
Russian 
German 
French 
Japanese 
Portuguese 
Hindi 