https://repositorio.ufu.br/handle/123456789/20866 2026-04-06T01:59:59Z https://repositorio.ufu.br/handle/123456789/48372 Title: Inteligência artificial para predição de diabetes mellitus tipo 2 com base em dados sociodemográficos e estilo de vida Abstract: Non-communicable chronic diseases, especially type 2 Diabetes Mellitus (T2DM), represent one of the greatest global public health challenges due to their high prevalence and their economic and social impact. In this context, this study applied Artificial Intelligence (AI) techniques, with an emphasis on Machine Learning (ML), to predict T2DM risk based on sociodemographic, clinical, and lifestyle data. Two public datasets from the Behavioral Risk Factor Surveillance System (BRFSS) were used, comprising 250,360 records and 21 variables. The J48 (C4.5) algorithm was implemented using Weka 3.8.6 software with 10-fold cross-validation. The model achieved an average accuracy of 83.85% for binary classification and 88.84% for multiclass classification. Feature selection identified six most relevant variables: hypertension, high cholesterol, heart disease, excessive alcohol consumption, self-rated health, and difficulty walking. The results demonstrate the potential of AI techniques for the early identification and prevention of T2DM, reinforcing the importance of integrating clinical, nutritional, and behavioral data in the development of predictive models. It is concluded that the application of AI in precision nutrition can optimize monitoring and preventive diagnosis, reducing costs and promoting quality of life. 2025-12-15T00:00:00Z https://repositorio.ufu.br/handle/123456789/48089 Title: Aplicação de algoritmos de machine learning para análise e predição de desfechos clínicos em pacientes com covid-19 Abstract: The COVID-19 pandemic imposed substantial challenges on healthcare systems worldwide, particularly due to the rapid clinical deterioration observed in many patients and the urgent need to allocate ventilatory support and intensive care resources efficiently, highlighting a gap in objective tools capable of assisting early risk stratification and outcome prediction. In this context, Machine Learning (ML) models, such as decision trees, offer the potential to support clinical decision-making by analyzing multiple variables simultaneously and generating easily interpretable predictive structures. This study aimed to apply ML algorithms to clinical data from hospitalized patients with COVID-19 to predict relevant clinical outcomes and identify factors associated with disease severity. This retrospective, quantitative study used secondary data from the public Severe Acute Respiratory Infection (SARI) database of Londrina, Paraná, Brazil, which initially contained 15,655 records of hospitalized patients between January 2021 and February 2022. After data cleaning and preprocessing, 5,704 records were used for the outcome prediction model and 7,182 for complementary analyses. Data were analyzed using the WEKA software (v.3.8.6) with the J48 decision tree algorithm and cross-validation, following attribute selection and removal of redundant or incomplete variables. The generated models achieved approximately 80% accuracy. Type of ventilatory support emerged as the most relevant predictor across analyses, followed by ICU admission, preexisting heart disease, age, and hospital type. Results showed that invasive mechanical ventilation and ICU admission were strongly associated with mortality, whereas non-invasive ventilation or absence of ventilatory support were associated with recovery. Advanced age substantially increased mortality risk among patients receiving invasive ventilation, and hospital type influenced outcomes among those not admitted to the ICU, with higher mortality in public hospitals. The study concludes that decision tree models are effective for identifying predictors of severity in hospitalized COVID-19 patients, providing clinically interpretable structures with potential application for risk stratification, resource allocation, and management of critical care pathways, especially in contexts of healthcare overload such as that experienced during the pandemic. 2025-12-18T00:00:00Z https://repositorio.ufu.br/handle/123456789/46518 Title: Estratégias para uso de genes inseticidas em Bacillus thuringiensis: expressão heteróloga e a análises genômicas Abstract: The demand for more sustainable methods of agricultural pest control has driven the development of bioinsecticides based on Bacillus thuringiensis (Bt), a bacterium capable of producing highly specific and biodegradable insecticidal proteins. However, the narrow spectrum of action of some toxins and the emergence of resistant insect populations highlight the need to expand the diversity of available insecticidal proteins. This dissertation evaluated complementary strategies to enhance the use of insecticidal genes in Bt by combining genomic analyses of strains for the discovery of new genes with genetic transformation aimed at heterologous expression of multiple toxins. Twenty-five Bt genomes available in the GenBank-NCBI database were analyzed, revealing that 52% of the strains carry known insecticidal genes, while 48% showed no identifiable genes, suggesting unexplored diversity. A total of 64 toxins were identified, mainly Cry proteins, with diverse chromosomal and plasmid distributions, highlighting promising strains such as IBL200, NB-176, BMP144, and C15 for the development of new bioinsecticides. In a second stage, genetic transformation of B. thuringiensis subsp. kurstaki HD-1 was performed using a polycistronic cassette containing the cyt2Aa, cry3Bb1, and cry9Ca1 genes, cloned into the pHT01 plasmid and propagated in Escherichia coli before transfer to Bt by electroporation. Protocols for competent cell preparation and electroporation were optimized, and gene integration was confirmed by PCR. The successful transformation of the HD-1 strain with multiple insecticidal genes represents an important step toward developing bioinsecticides with an expanded spectrum of action, supporting integrated pest management for tomato crops with reduced reliance on chemical insecticides. 2025-06-06T00:00:00Z https://repositorio.ufu.br/handle/123456789/46474 Title: Reposicionamento de fármacos por docking molecular direcionado às proteínas PNMA3 e RASSF2 associadas à resistência à quimioradioterapia no câncer de colo uterino Abstract: Cervical cancer (CC) remains one of the most lethal malignancies among women, particularly in low- and middle-income countries. Its primary etiology is associated with persistent infection by oncogenic human papillomavirus (HPV) genotypes. Worsening this scenario, chemoradiotherapy resistance in certain patients remains a significant clinical challenge. Previous transcriptomic studies identified the overexpression of RASSF2 and PNMA3 genes in non-responsive patients, raising the hypothesis that PNMA3 might disrupt the association between RASSF2 and MST1/MST2 kinases, which are critical components of the Hippo apoptotic pathway. This study aimed to investigate, through structural bioinformatics approaches, whether PNMA3 and RASSF2 interact, and if so, to identify the interaction site and assess its impact on Hippo pathway activation. Rational drug repositioning was also explored as a strategy to modulate or block this interaction. Three-dimensional models of PNMA3 and RASSF2 were obtained from public databases, and a drug library was generated through systematic filtering of the ChEMBL database. Protein–protein and protein–ligand docking analyses, followed by molecular dynamics simulations, were performed using the MOE software, with subsequent data processing in R. The results indicated that PNMA3 forms a strong and stable interaction with RASSF2 at the SARAH domain, potentially impairing apoptotic signaling. Virtual screening revealed compounds with differential selectivity. Among the candidates, Nebivolol stood out as the most promising, exhibiting high affinity for PNMA3 (S = –9.42 kcal/mol) and no significant interaction with RASSF2, supporting its potential to selectively disrupt this pathological interaction. In this context, rational drug repositioning emerges as a promising strategy to design alternative therapeutic approaches to overcome resistance in cervical cancer. Notes: Este trabalho contou com o apoio da Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), do Ministério da Saúde do Brasil por meio do Programa Nacional de Apoio à Atenção Oncológica – Pronon (NUP: 25000.159953/2014-18 e NUP: 25000.079266/2015-09), da Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG; APQ-02255-22) e da Rede Mineira de Pesquisa Translacional em Oncologia (RED 00059-23). 2025-07-04T00:00:00Z