COMPUTER-ASSISTED SIMULATION OF C-13 NUCLEAR MAGNETIC SPECTRA OF MONOSACCHARIDES

Mathematical models are developed that relate the structures of monosa ccharides to their C-13 nuclear magnetic resonance spectra. The data s et of monosaccharides consists of 55 monosaccharides in the six-member ed ring configuration and 56 monosaccharides in the five-membered ring configuration. The structural environment of each carbon atom in the data set is encoded using numerical atom-based descriptors which are t hen used to develop linear regression models relating the C-13 chemica l shift to the structural features. The atom-based descriptors used in this study encode topological, geometric, and electronic information about the carbon atoms in monosaccharides. Multiple linear regression analysis is used to develop an 11-descriptor model to predict the chem ical shifts of pyranoses and pyranosides and an eight-descriptor model to predict the chemical shifts of furanoses and furanosides. The mode ls are then submitted to computational neural networks, giving improve d results with final training set rms errors of 1.03 ppm for pyranoses and pyranosides and 1.58 ppm for furanoses and furanosides.