Automatic compatibility tests of HSQC NMR spectra with proposed structures of chemical compounds

A recently introduced similarity measure is extended here for comparing two-dimensional spectra. Its applicability is demonstrated with heteronuclear single-quantum correlation (HSQC) NMR spectra. For testing the compatibility of a spectrum with the proposed chemical structure, first, the spectrum is predicted on the basis of that structure and then, the proposed comparison algorithm is applied. In this context, the topics of optimization are peak picking, signal intensity measures, and optimizing the parameters of the two-dimensional comparison method. The performance is analyzed with a test set of 289 structures of organic compounds and their HSQC and 1H NMR spectra. The results obtained with HSQC spectra are better than those achieved using the previously described one-dimensional similarity test with 1H NMR spectra alone.