{"title":"BuildQSAR: A New Computer Program for QSAR Analysis","authors":"Daniel Oliveira, A. C. Gaudio","doi":"10.1002/1521-3838(200012)19:6<599::AID-QSAR599>3.0.CO;2-B","DOIUrl":null,"url":null,"abstract":"A new computer program called BuildQSAR has been designed to help the QSAR practitioner on the task of building and analyzing quantitative models through regression analysis. The main part of the program is a spreadsheet, in which the user can enter with the data set composed by the structure definition of the compounds, one or more types of biological activity values and many physicochemical properties. The program has an external data bank, which includes the values of many known substituent parameters. The common tasks that can be performed by the program are: (a) linear, quadratic and bilinear models; (b) graphics analysis; (c) correlation analysis and; (d) hypothesis testing. BuildQSAR can also: (e) build subsets of the data set that include groups of compounds chosen by the user; (f) create new variables, derived from the existing ones, and create many types of indicator variables; (g) build data sets directly from the data bank by selecting the substituents and the variables therein; (h) perform variable selection by systematic search and genetic algorithm; (i) perform cross-validation of specific models; (j) eliminate outliers and analyze the resulting model; and others. The program was coded in Delphi 5 and runs in Windows 95/98/00/NT environments.","PeriodicalId":20818,"journal":{"name":"Quantitative Structure-activity Relationships","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2000-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"159","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Quantitative Structure-activity Relationships","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1002/1521-3838(200012)19:6<599::AID-QSAR599>3.0.CO;2-B","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 159
Abstract
A new computer program called BuildQSAR has been designed to help the QSAR practitioner on the task of building and analyzing quantitative models through regression analysis. The main part of the program is a spreadsheet, in which the user can enter with the data set composed by the structure definition of the compounds, one or more types of biological activity values and many physicochemical properties. The program has an external data bank, which includes the values of many known substituent parameters. The common tasks that can be performed by the program are: (a) linear, quadratic and bilinear models; (b) graphics analysis; (c) correlation analysis and; (d) hypothesis testing. BuildQSAR can also: (e) build subsets of the data set that include groups of compounds chosen by the user; (f) create new variables, derived from the existing ones, and create many types of indicator variables; (g) build data sets directly from the data bank by selecting the substituents and the variables therein; (h) perform variable selection by systematic search and genetic algorithm; (i) perform cross-validation of specific models; (j) eliminate outliers and analyze the resulting model; and others. The program was coded in Delphi 5 and runs in Windows 95/98/00/NT environments.