Pharmacophore identification and bioactivity prediction for group I metabotropic glutamate receptor agonists by the electron-conformational QSAR method
{"title":"Pharmacophore identification and bioactivity prediction for group I metabotropic glutamate receptor agonists by the electron-conformational QSAR method","authors":"E. Rosines, I. Bersuker, J. E. Boggs","doi":"10.1002/1521-3838(200111)20:4<327::AID-QSAR327>3.0.CO;2-Q","DOIUrl":null,"url":null,"abstract":"The pharmacophore for group I metabotropic glutamate receptor (mGluRl) agonists is revealed and their activity predicted by means of the previously developed and further improved electron-conformational (EC) method. A distinguishing feature of this method is that in addition to revealing the pharmacophore of activity as a set of specific atomic electronic features arranged in a special geometry, it allows for prediction of the activity quantitatively as a function of the parameters of pharmacophore flexibility and anti-pharmacophore shielding groups. Conformational analysis, electronic structure calculations, and matrix processing are performed for the training set of 29 compounds, 13 active and 16 inactive, and the pharmacophore of mGluRl agonists is evaluated. It contains a four-point skeleton of three oxygen atoms and one nitrogen atom at certain interatomic distances with restricted atomic interaction indices whereby all these parameters are determined within certain tolerances. The pharmacophore parameter flexibilities, as well as the influence of the anti-pharmacophore shielding and other auxiliary groups are parameterized and weighted by seven constants, their values being obtained from a least-square regression with very good statistics: R 2 =0.97, F=589 (∼100% level of confidence), and a standard error of about 5% of the range of measured values. The results are also tested with the leave-one-out cross-validation method that yields prediction statistics R 2 = 0.91. The E statistics were also evaluated illustrating the role of each of the activity parameters involved.","PeriodicalId":20818,"journal":{"name":"Quantitative Structure-activity Relationships","volume":"13 1","pages":"327-334"},"PeriodicalIF":0.0000,"publicationDate":"2001-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"7","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Quantitative Structure-activity Relationships","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1002/1521-3838(200111)20:4<327::AID-QSAR327>3.0.CO;2-Q","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 7
Abstract
The pharmacophore for group I metabotropic glutamate receptor (mGluRl) agonists is revealed and their activity predicted by means of the previously developed and further improved electron-conformational (EC) method. A distinguishing feature of this method is that in addition to revealing the pharmacophore of activity as a set of specific atomic electronic features arranged in a special geometry, it allows for prediction of the activity quantitatively as a function of the parameters of pharmacophore flexibility and anti-pharmacophore shielding groups. Conformational analysis, electronic structure calculations, and matrix processing are performed for the training set of 29 compounds, 13 active and 16 inactive, and the pharmacophore of mGluRl agonists is evaluated. It contains a four-point skeleton of three oxygen atoms and one nitrogen atom at certain interatomic distances with restricted atomic interaction indices whereby all these parameters are determined within certain tolerances. The pharmacophore parameter flexibilities, as well as the influence of the anti-pharmacophore shielding and other auxiliary groups are parameterized and weighted by seven constants, their values being obtained from a least-square regression with very good statistics: R 2 =0.97, F=589 (∼100% level of confidence), and a standard error of about 5% of the range of measured values. The results are also tested with the leave-one-out cross-validation method that yields prediction statistics R 2 = 0.91. The E statistics were also evaluated illustrating the role of each of the activity parameters involved.