Taxonomically informed metabolite annotation and data organization in natural products research

Natural products, more precisely defined as specialized metabolites, are by definition strongly linked to the taxonomical position of the producing organisms. Considering taxonomy when exploring natural products thus appears as an evidence. Such principles were already formulated in 1816 by De Candolle who postulated that 1) Plant taxonomy would be the most useful guide to man in his search for new industrial and medicinal plants and 2) Chemical characteristics of plants will be most valuable to plant taxonomy in the future. [1] We adhere to De Candolle’s postulate and aim to establish their validity using computational approaches.

Regarding the first postulate, we show that the consideration of taxonomic position is beneficial in the metabolite annotation process leading to a systematic improvement (> 50% of correct annotation at rank 1) of current in silico metabolite annotation solution (Sirius, MSFinder, CFM+Tremolo). [2] This increased confidence in metabolite annotation efficiently improves the natural products drug discovery process and should complement orthogonal information already shown to strengthen classical spectral scoring systems. [3],[4] Regarding the second postulate, we demonstrate the interest of considering the chemical dimension (structural or spectral relationships) when seeking to organize plants or lichens extracts. We present a novel metric, Spectral and Substructural Similarity- Informed Distance (SSS-ID) for such classification and compare it to classically established taxonomies (genetic material or morphological character-based organizations), demonstrating its interest. The principles and implementation as well as practical applications of such approaches will be discussed.

• References

• 1 Candolle AP de. Essai sur les propriétées médicales des plantes, comparées avec leurs formes extérieures et leur classification naturelle. Paris: Chez Crochard, Libraire,; 1816. On Internet: http://www.biodiversitylibrary.org/bibliography/112422
• 2 Dounoue-Kubo M, Rutz A, Bisson J, Saesong T, Bagherri M, Ebrahimi SN. et al. Taxonomically informed scoring enhances confidence in natural products annotation. Front Plant Sci. 2019 (in preparation).
• 3 da Silva RR, Wang M, Nothias LF, JJJ van der Hooft, Caraballo-Rodríguez AM, Fox E. et al. Propagating annotations of molecular networks using in silico fragmentation. PLoS Comput Biol 2018; 14: 1-26.
• 4 Bach E, Szedmak S, Brouard C, Böcker S, Rousu J. Liquid-chromatography retention order prediction for metabolite identification. Bioinf 2018; 34: i875-i883. On Internet: https://academic.oup.com/bioinformatics/article/34/17/i875/5093227