New Dereplication Method Applied to NMR-Based Metabolomics on Different Fusarium Species Isolated from Rhizosphere of Senna spectabilis

The search for new sources of natural products steadily increased the use of bioinformatic tools that enabled efficient analysis of complex matrices. In this context, dereplication methods emerged as a fast way of identifying known compounds, accelerating the identification of bioactive chemotypes. Although 1H NMR is widely used as an analytical technique, few studies have been reported using it as a dereplication tool, primarily because of the spectral complexity. This work aims to create a new computational method that analyses 1H NMR data from Fusarium solani and F. oxysporum isolated from Senna spectabilis's rhizosphere through principal component analysis (PCA). The algorithm uses loading values to select important peaks that distinguish both species in PCA, allowing compound dereplication, even in highly similar profiles. As a result, the method, associated with other NMR experiments and information from an in-house Fusarium's metabolite library was able to distinguish different mycotoxins produced by both fungi, identifying fusaric acid and beauvericin for F. oxysporum and the depsipeptide HA23 from F. solani.