Theoretical NMR correlations based Structure Discussion

Abstract

The constitutional assignment of natural products by NMR spectroscopy is usually based on 2D NMR experiments like COSY, HSQC, and HMBC. The actual difficulty of the structure elucidation problem depends more on the type of the investigated molecule than on its size. The moment HMBC data is involved in the process or a large number of heteroatoms is present, a possibility of multiple solutions fitting the same data set exists. A structure elucidation software can be used to find such alternative constitutional assignments and help in the discussion in order to find the correct solution. But this is rarely done. This article describes the use of theoretical NMR correlation data in the structure elucidation process with WEBCOCON, not for the initial constitutional assignments, but to define how well a suggested molecule could have been described by NMR correlation data. The results of this analysis can be used to decide on further steps needed to assure the correctness of the structural assignment. As first step the analysis of the deviation of carbon chemical shifts is performed, comparing chemical shifts predicted for each possible solution with the experimental data. The application of this technique to three well known compounds is shown. Using NMR correlation data alone for the description of the constitutions is not always enough, even when including 13C chemical shift prediction.

Incorporation of 4J-HMBC and NOE Data into Computer-Assisted Structure Elucidation with WebCocon

“…In order to address this question, WebCocon can generate a complete theoretical NMR correlation data set (COSY, HMBC, NHMBC, and ADEQ data) for a molecule. These data can then be submitted to the WebCocon server for a structure elucidation [ 32 ].…”

“…Over the decades, many different methods have been implemented, the most prominent being fragment assemblers [ 1 , 2 , 3 , 4 , 5 , 6 ], expert systems [ 7 , 8 , 9 ], structure generation by reduction [ 10 ], logic engines [ 11 ], stochastic structure generators [ 12 ], combinatorial brute force [ 13 , 14 , 15 , 16 , 17 ], databases of C NMR chemical shifts and fragments [ 18 , 19 ], combinatorial structure generation with restraints [ 20 , 21 ], genetic algorithms [ 22 , 23 ], simulated annealing [ 24 ], convergent structure generation [ 25 , 26 ], evolutionary algorithm [ 27 ], fuzzy structure generation [ 28 ], and expert systems with DFT [ 29 ]. However, CASE remains a challenge [ 29 , 30 , 31 , 32 , 33 , 34 ]. The basic issue is that the relation between a small molecule and its NMR correlation data is not reciprocal.…”

Concurrent combined verification: reducing false positives in automated NMR structure verification through the evaluation of multiple challenge control structures

“…Ideally, an automated structure generation method would use some type of statistical filtering of observed NMR data as reported elsewhere. [6,7] By taking into account some aspect of the observed NMR data, this would help to produce more relevant challenge structures for the highest quality evaluation. For the bulk of the analysis, two challenge structures were used.…”

Surge - A Fast Open-Source Chemical Graph Generator

“…Many works followed but few examples of practical usability are available even today [12]. Among the currently available structure generators, such as DENDRAL, ASSEMBLE, SMOG, COCON [13] and LSD [14], MOLGEN [15] constituted the state-of-the-art for decades in terms of speed, completeness and reliability. The first version of MOLGEN was based on the strategy of DENDRAL software and developed to overcome the limitations of DENDRAL [16].…”