Role of in silico tools and text mining in the safety assessment of selected plant coumarins

Giuseppa Raitano, Emilio Benfenati, Sophia Ananiadou, Xiao Fu, Quoc Tuan Do, Patrice Rat, Laurent Sousselier

Research output: Contribution to journalArticlepeer-review


Objective: Plant extracts are widely used in a number of industrial sectors. Their complexity and, for cosmetics, that animal tests are no longer possible, warrants a new approach based on natural molecules to allow for their safety assessment. Using coumarins as an example, this study outlines new methodologies developed in the frame of the NCSTOX project to undertake safety assessments of plant components based on in silico tools and text mining.

Methods: Instead of assessing each extract, information was gathered on all plant constituents. Coumarins, which are present in numerous plants and known to exhibit safety concerns, were chosen to validate the methodology. An innovative multi-step text mining approach, using the integrative ARGO workbench, was combined with the use of various in silico models, including VEGA, to predict safe levels of use. Critical compounds were identified as those associated with genotoxicity or classified as high potency skin sensitizers. Other molecules were classified according to a Threshold of Toxicological Concern (TTC) approach.

Results: Information on 200 coumarins was compiled. The results demonstrate the key modulating effect and role of the position of certain substituents (e.g. the hydroxyl group) on the coumarins’ scaffold with regard to their safe level of use.

Conclusions: The safety assessment of selected coumarins validated a new methodology to establish the toxicological profile of plant constituents. Using text mining methods we curated a novel database to provide the scientific community with animal-free, safe levels of use for plant constituents.
Original languageEnglish
Pages (from-to)S96
JournalToxicology Letters
Early online date2 Nov 2017
Publication statusPublished - 2017


Dive into the research topics of 'Role of in silico tools and text mining in the safety assessment of selected plant coumarins'. Together they form a unique fingerprint.

Cite this