Desenvolvimento de plataforma biofotônica associada a algoritmos de inteligência artificial para identificação de SARS-CoV-2 em saliva artificial

• Autor(a) principal: Taveira, Elisa Borges
• Data de Publicação: 2022
• Tipo de documento: Trabalho de conclusão de curso
• Idioma: por
• Título da fonte: Repositório Institucional da UFU
• Texto Completo: https://repositorio.ufu.br/handle/123456789/35351
• topic: ATR-FTIR; Covid-19; SARS-CoV-2; Teste diagnóstico; Diagnóstico salivar; ATR-FTIR; Covid-19; SARS-CoV-2; Diagnostic-test; Salivary-diagnosis; CNPQ::CIENCIAS BIOLOGICAS::FISIOLOGIA

The rapid spread of COVID-19 and SARS-CoV-2 variants represents a global issue, even after vaccine development. The development of novel sustainable screening platforms to detect SARS-CoV-2 directly is a faster alternative to serologically-based assays. Here, we used attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR) supported by artificial intelligence algorithms to identify unique infrared vibrational modes of a pseudotyped human immunodeficiency virus type-1 (HIV-1) coupled to Spike (S) protein of SARS-CoV-2 (HIV/NanoLuc-SARS-CoV-2 pseudotype virus), diluted in eight different concentrations in artificial saliva samples to develop a self-collected, reagent-free, and green technology for SARS-CoV-2 detection. Linear Discriminant Analysis algorithm was applied to differentiate 2.27 x 106 RLU/ml, 1.14 x 106 RLU/ml, 5.68 x 105 RLU/ml, 2.84 x 105 RLU/ml, 1.42 x 105 RLU/ml, 7.10 x 104 RLU/ml, 3.55 x 104 RLU/ml, and 1.77 x 104 RLU/ml of HIV/NanoLuc-SARS-CoV-2 in artificial saliva. The discrimination for four different concentrations in 1.14 x 106 RLU/ml, 5.68 x 105 RLU/ml, 2.84 x 105 RLU/ml and 1.42 x 105 RLU/ml varied between 88% and 94%. The performance for 7.10 x 104 RLU/ml was 85%. The discrimination for the lower concentrations in 3.55 x 104 RLU/ml and 1.77 x 104 RLU/ml were 72% and 79%, respectively. Our findings demonstrate a potential application of this reagent-free biophotonic platform supported with machine learning algorithms to detect SARS-CoV-2 in saliva.