No abstract

The SPECTRA collection of software as a spectral information management system for organic compound structure determination is described. The SPECTRA (SPECTral Research and Analysis) system suggests candidate structures for chemical compounds based on analysis of their spectra, where mass spectra, infrared spectra, 1H-nuclear magnetic resonance spectra, and 13C-nuclear magnetic resonance spectra are possible input. The system computes the optimal matching of an input spectrum with stored spectra in a database and also retrieves the spectra of compounds that contain a substructure of the unknown compound. A novel combined search algorithm can be activated when two to four spectra are given as information of an unknown compound. Similarities between the input spectrum and each spectrum in the database are calculated, and the corresponding candidate compounds are ranked according to their similarity score.

Computer programs called PREHAC1 and PREHAC2, which aid in the impartial selection of substituents for the synthesis of highly bioactive compounds among congeners based on the Hansch‐Fujita analysis, were developed. The PREHAC1 and 2 programs are for aliphatic and aromatic monosubstituted derivatives and for aromatic disubstituted derivatives, respectively. Thirteen sets of physicochemical parameter values of 408 useful substituents were previously input into a data file as Master Data. PREHAC programs are activated by the input of the coefficients and intercept of the QSAR equation with the corresponding parameter codes, the information on the number of the substituents to be printed out, etc., and then Master Data is searched. Calculated activity values are ranked in descending order and the high‐ranking substituents are printed out according to the specified number. An example is presented to illustrate the programs.

Application of a new chemometric system, SPECTRE, to quantitative structureactivity relationship (QSAR) analysis in agricultural drug design has been studied. The SPECTRE system was employed to analyze experimental data by calculating a statistical predictive model using an evolution of the PLS (Partial Least Squares) regression method. This new modeling method, which does not need any a priori knowledge about the chemistry involved, is compared with multi-linear regression (MLR) analysis where the performance depends upon knowledge provided by the researcher. The SPECTRE system is shown to be able to produce similar or even superior results when compared with the conventional method.

No abstract