Computer-assisted structural analysis of regular glycopolymers on the basis of 13C NMR data.

A computer-assisted approach to the prediction of the primary structures of regular glycopolymers is described. The analysis is based on comparing the calculated 13C NMR spectra of all the possible structures of the repeating unit (for the given monomeric composition) to an experimental 13C NMR spectrum. The spectra generation is based on the spectral database containing information on the 13C chemical shifts of monomers, di- and trimeric fragments. If the required data are missing from this database, the special database for average glycosylation effects is used. The analysis reveals those structures with the calculated 13C NMR spectrum most close to observed. The structures of repeating units of any topology containing up to six residues linked by glycosidic, amidic or phospho-diester bridges can be predicted. Unambiguous selection of the proper structure from the output list of possible structures may require additional experimental data. Testing the created program and databases on bacterial polysaccharides and their derivatives containing up to three non-sugar residues (alditols, amino acids, phosphate groups etc.) per repeating unit revealed the good convergence of prediction with independently obtained structural data.