Structure elucidation of organic natural products is a complex task mainly performed
using spectroscopic methods. Among the techniques applied, NMR-spectroscopy plays
a central role giving a detailed insight into constitution, configuration and conformation
of the unknown compound. Despite the tremendous progress made in tailoring sophisticated
pulse techniques to solve structural problems, many natural products have been published
showing a wrong structure proposal leading to incorrect reference material for subsequent
conclusions with the effect of making this erroneous knowledge base statistically
more significant. A large number of comprehensive review articles [1],[2] describes the actual situation in detail. From this analysis the necessity of massive
software application during every step of the process of structure elucidation can
be derived. The automatic peer-reviewing of some 700,000 C13-NMR spectra mainly taken
from the public domain literature has revealed a significant number of wrong structure
proposals and/or wrong signal assignments [3]. Subsequent application of a structure generator program based on the carbon-NMR
data has created reasonable alternatives – in many cases in full agreement with an
already known structure revision. Examples will be given showing the immediate need
to change the workflow of structure elucidation along the line from spectrum interpretation
to peer-reviewing of manuscripts [4].
