Espectroscopia Raman (Portátil) e quimiometria para distinção entre uma cultivar de algodão convencional e uma geneticamente modificada

• Autor(a) principal: Mata, Mayara Macedo da
• Data de Publicação: 2021
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da UEPB
• Texto Completo: http://tede.bc.uepb.edu.br/jspui/handle/tede/3822
• topic: Cultura do algodão; Transgênicos; Quimiometria; CIENCIAS EXATAS E DA TERRA::QUIMICA

Transgenics undergo genetic changes in order to exploit superior agronomic traits through genetic engineering methods. Biotechnological techniques are used for the improvement of species as well as to solve problems related to quality, productivity and resistance to pests. Cotton as one of the main commodities for the country has a high technological level that guarantees it is produced with environmental sustainability (the smallest water footprint in the world), social sustainability, and fiber quality. At the same time, Brazil already stands out with great potential for organic cotton in semi-arid areas, involving small producers. As a rationale, society's new consumption patterns require sustainable production guarantees, and this requires as an essential condition the use of exploratory, rapid, and economically feasible analytical methods. Vibrational spectroscopy has a set of techniques with attractive characteristics for developing methods to meet these technological and commercial requirements. Among them, Raman spectroscopy as analytical instrumentation and explored using multivariate models is an alternative that has advantages for implementation studies. The objective of this work was to develop methods for distinguishing seeds in conventional and transgenic cotton cultivars using Raman spectroscopy and pattern recognition techniques. Twenty individual cotton seed samples (10 conventional and 10 transgenic) were used. Raman spectra of all samples were recorded and a PLS-DA model was constructed to distinguish between cottonseeds. In the training stage 97.1% of hits and an RMSE of 0.029 were obtained, and in the test stage 96.7% of hits and an RMSE of 0.033 were obtained. Distinguishing between conventional and transgenic cotton cultivars by means of pattern recognition models together with portable Raman spectroscopy is an alternative as a fast, easily implemented and cost-effective diagnostic method that can be explored and validated for large-scale implementation.