DENGUE VIRUS DETECTION BY MIDDLE INFRARED SPECTROSCOPY IN CONJUNCTION WITH GENETIC ALGORITHM AND LINEAR DISCRIMINANT ANALYSIS: A CHEMOMETRIC MODEL

• Autor(a) principal: Silva , Gigliane Joice Santos da
• Data de Publicação: 2021
• Outros Autores: Filho , Marcos Claudino Batista dos Santos, Santos, Paulo Venicius Messias dos, Santos, Ricardo Fernandes dos, Santos, Marfran Claudino Domingos dos
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Scientia Generalis
• Texto Completo: https://scientiageneralis.com.br/index.php/SG/article/view/161
• topic: Dengue detection; Computational algorithms; Pattern Recognition; GA-LDA; Detecção da Dengue; Algoritmos computacionais; Reconhecimento de padrões; GA-LDA

The aim of this study was to develop a chemometric model based on the application of the GA-LDA computational algorithm in the analysis of mid-infrared data to detect dengue virus from a set of spectra of clinical samples. In this approach, the GA selects the most important variables in the set of spectra, while the LDA. works on the discrimination between classes, based on these selected variables. The study investigated the advantages of this approach compared to standard techniques, since there is a need for a technique that allows for a quick response with low cost and good specificity. The data set used in this study was cordially made available by the Research Group on Biological Chemistry and Chemometrics at UFRN. The model was built using the “MATLAB” software, where computational analysis and calculations of quality measures (figures of merit) were performed. It was found that the technique achieved a good prediction, with 100% sensitivity and specificity, of the correct class of all samples used in the test set. Furthermore, it has been shown not to need the use of reagents or kits for the analysis, in addition to providing faster results compared to the standard techniques used. From these results, it was possible to verify the importance of applying a computational tool in the field of virology, since a great potential for this application was observed. However, further studies are needed.