2025-02-102025-04-122023-11-01ALCÂNTARA, Bianca Gonçalves Vasconcelos de. Metabolômica, Fitoquímica e Avaliação do Potencial An--inflamatório de Espécies de Lauraceae. 2023. 125 f. Tese (Doutorado em Química) - Universidade Federal de Alfenas, Alfenas, MG, 2023 .https://repositorio.unifal-mg.edu.br/handle/123456789/2510The Lauraceae family has a great diversity of genera and species, of which several species are endemic to Brazil. Some species of this family have demonstrated anti-inflammatory properties for their extracts or isolated substances. Inflammation is a disorder present in a wide range of pathologies that represents a relevant target for studies to discover new treatments. In this context, this work aimed to evaluate the anti-inflammatory activity for species belonging to the Lauraceae family. Furthermore, carry out a metabolomic study of these species, obtain an insilico prediction model of anti-inflammatory activity and select a species to carry out a phytochemical study. UPLC-HRMS was used for the metabolomics study. An ex vivo model was used for anti-inflammatory evaluation. In addition to statistical tools to obtain a model for predicting the anti-inflammatory activity of plant extracts from the Lauraceae family. Using metabolomic strategies, it was possible to direct the isolation and purification of a biomarker of anti-inflammatory activity. The ex vivo results demonstrated that most species presented anti-inflammatory activity greater than 50% inhibition, with the species Ocotea diospirifolia, Cinnamomum glaziovii, Nectandra membranacea, Ocotea odorifera and Persea major presenting the highest values of percentage of PGE2 inhibition, being above 77% inhibition. Several extracts demonstrated anti-inflammatory activity for the first time. Furthermore, the metabolomic study allowed us to chemically characterize the samples and build a robust PLS prediction model, with values of R 2 = 0.99 and Q 2 = 0.92 and values predicted in external validation very close to the real values. Furthermore, additional ANN and SGD machine learning models proved to be robust and corroborated the results of the PLS model. Among the biomarkers positively correlated with anti-inflammatory activity, aporphinic and benzylisoquinoline alkaloids were noted with confidence level 2, according to metabolomic guidelines. Furthermore, it was possible to isolate one of the biomarkers of anti-inflammatory activity, the new alkaloid 2,10-dimethoxyaporphine, from the specie O. odorifera. Thus, this work in general produced new knowledge about 18 species of the Lauraceae family that could serve as a basis for future studies of species of the Lauraceae family; in addition to highlighting the potential for discovering new chemical compounds.application/pdfAcesso EmbargadoPrediçãoInflamaçãoAnálises de correlação in silicoAlcaloidesQUIMICA::QUIMICA ORGANICATesePaula, Daniela Aparecida Chagas De