NEW APPROACHES FOR STRUCTURAL ELUCIDATION USING NMR CALCULATIONS

Nuclear magnetic resonance (NMR) is undoubtedly the most important spectroscopic technique for the structural and stereochemical elucidation of natural products. However, despite the impressive advances that have been made in this field, it is striking the large number of structures erroneously assigned based on spectroscopic data [1].
Quantum chemical calculations of NMR shifts and coupling constants emerge as a powerful and simple option to shed light on structural or stereochemical issues of complex natural products. This approach has been extensively employed in recent years to facilitate the determination of the tridimensional structure of a wide variety of organic molecules [2].
Our research team has made important contributions in this field [3]. We pioneered the use of artificial intelligence methods in the development of novel strategies of structural validation (ANN-PRA) [4a]. In a conceptually different approach, we developed DP4+, a probability used to determine the most likely structure among two or more candidates when one set of experimental data is available [4b]. Despite the results obtained with DP4+ are generally excellent, one of its limitations is related to the relatively large computational cost involved in obtaining DFT-optimized geometries. This led to the development of a faster variant including the valuable information provided by coupling constants (J-DP4 method) [4c]. However, there remain considerable challenges, such as the case of configurational assessment of polar systems featuring multiple intramolecular hydrogen bonding interactions because of the poor energy predictions provided by most DFT methods. In our latest work, we tackle this problem by averaging the results provided by randomly generated ensembles, paving the way for a new paradigm in quantum NMR-assisted structural elucidation [4d]. These tools were explored to suggest the most probable structure of controversial natural or unnatural products originally misassigned, with some predictions further validated by synthesis.