Galamba, Nuno Jorge Rosa Lopes, 1974-Martiniano, Hugo Filipe de Mesquita Costa, 1978-Travessa, Pedro Miguel Carvalho2025-01-132025-01-1320242024http://hdl.handle.net/10400.5/97107Tese de mestrado, Bioinformática e Biologia Computacional, 2024, Universidade de Lisboa, Faculdade de CiênciasNeurodegenerative diseases like Alzheimer’s and Parkinson’s present significant challenges in modern medicine due to the progressive degeneration of neuronal cells, leading to severe cognitive and motor impairments. A key feature of these disorders is protein aggregation, which disrupts cellular functions and causes neuronal death. In Parkinson’s disease (PD), the aggregation of alpha-synuclein (α-syn), and in Alzheimer’s, amyloid beta (Aβ) and tau, are critical molecular events. Understanding the structural and dynamic changes driving protein aggregation is crucial for developing therapeutic interventions. These diseases share aberrant protein conformations that promote seed aggregation, with misfolded proteins forming toxic aggregates that contribute to neuronal dysfunction and cell death. In this study, we utilized molecular simulations and structural descriptors to investigate α-syn behavior and identify potential drugs. We combined molecular dynamics (MD) simulations with the ProtT5 model to explore the ability of cyclic peptides to bind to specific regions of α-syn and their impact on the protein’s conformational space. Our approach involved two key steps: first, using MD simulations to examine α-syn conformational changes in the presence of cyclic peptides, and second, employing ProtT5 to generate embeddings for efficiently estimating the radius of gyration (Rg) values of α-syn with new peptides. The random forest regression model predicted Rg values, reducing the need for costly molecular simulations. Additionally, we developed a Python script for peptide selection in future investigations. Although this tool shows promise in the peptide selection process, it requires further refinement due to the small dataset size, which currently limits the robustness of predictions. This study improves our understanding of protein-cyclic peptide interactions, potentially leading to new treatment options that target α-syn aggregation pathways. Our results help to advance the greater objective of identifying effective therapies for neurodegenerative diseases, using fast and cheaper in silico methods.engAlterações Conformacionais de Proteínasα-synSimulações de Dinâmica MolecularAlgoritmos de aprendizagem automáticaInteligência Artificial (IA)Teses de mestrado - 2024master thesis203878329