Desenvolvimento de novas moléculas bioativas baseadas em componentes salivares, bioinformática e inteligência artificial: Estratégias de diagnóstico salivar e atividade antiviral para Mpox, Norovirus e o SARS-CoV-2

Abstract

This thesis highlights significant advancements in salivary detection of Mpox, Norovirus, and SARS-CoV-2, as well as the potential development of innovative bioactive antiviral molecules based on in silico docking of salivary peptides and artificial intelligence. The specific objectives include analyzing Mpox replication, its potential oral transmission, and the prospects of applying omics technologies in the salivary diagnosis of Mpox; developing a portable electrochemical device, modified with natural salivary peptides and coupled with artificial intelligence algorithms, for sensitive and specific detection of norovirus; creating and testing a portable electrochemical device with machine learning algorithms and artificial intelligence-generated peptides for the detection of SARS-CoV-2; and exploring natural and artificial salivary peptides as antiviral agents, evaluating their ability to inhibit viral entry and replication in host cells. Initially, we discuss the potential of using saliva as a diagnostic tool for mpox, emphasizing its presence in oral mucosal lesions and body fluids, and suggesting the application of omics technologies for non-invasive detection. In a study focused on norovirus, we used bioinformatics and AI to select salivary peptides (NVp1, NVp2, NVp3) that improved the electrochemical detection of the virus with high precision, sensitivity, and specificity. Another study targeted SARS-CoV-2, where AI-modeled peptides were used in an electrochemical biosensor, achieving 100% sensitivity, 80% specificity, and 90% accuracy for salivary detection of COVID-19, validated by machine learning algorithms. Additionally, we explored salivary peptides to block the interaction of the SARS-CoV-2 Spike protein with ACE2 receptors, demonstrating significant antiviral activity and reduced viral replication in variants. Finally, AI-generated peptides, BIAI1, BIAI2, and BIAI3, showed high antiviral efficacy against SARS-CoV-2 variants, significantly reducing viral titers and cell death, indicating their potential for prophylactic, diagnostic, and therapeutic applications. Collectively, these studies highlight the promise of integrating bioinformatics, AI, and salivary diagnostics for rapid, accurate, and non-invasive viral detection, as well as intervention strategies.