Publication
Geometric modeling, simulation, and visualization methods for plasmid DNA molecules
| datacite.subject.fos | Engenharia e Tecnologia::Engenharia Eletrotécnica, Eletrónica e Informática | pt_PT |
| dc.contributor.advisor | Gomes, Abel João Padrão | |
| dc.contributor.author | Raposo, Adriano Nunes | |
| dc.date.accessioned | 2016-02-25T15:25:24Z | |
| dc.date.available | 2016-02-25T15:25:24Z | |
| dc.date.issued | 2015-03 | |
| dc.description.abstract | Plasmid DNA molecules are a special type of DNA molecules that are used, among other applications, in DNA vaccination and gene therapy. These molecules are characterized by, when in their natural state, presenting a closed-circular conformation and by being supercoiled. The production of plasmid DNA using bacteria as hosts implies a purification step where the plasmid DNA molecules are separated from the DNA of the host and other contaminants. This purification process, and all the physical and chemical variations involved, such as temperature changes, may affect the plasmid DNA molecules conformation by uncoiling or even by open them, which makes them useless for therapeutic applications. Because of that, researchers are always searching for new purification techniques that maximize the amount of supercoiled plasmid DNA that is produced. Computer simulations and 3D visualization of plasmid DNA can bring many advantages because they allow researchers to actually see what can happen to the molecules under certain conditions. In this sense, it was necessary to develop reliable and accurate geometric models specific for plasmid DNA simulations. This dissertation presents a new assembling algorithm for B-DNA specifically developed for plasmid DNA assembling. This new assembling algorithm is completely adaptive in the sense that it allows researchers to assemble any plasmid DNA base-pair sequence along any arbitrary conformation that fits the length of the plasmid DNA molecule. This is specially suitable for plasmid DNA simulations, where conformations are generated by simulation procedures and there is the need to assemble the given base-pair sequence over that conformation, what can not be done by conventional predictive DNA assembling methods. Unlike traditional molecular visualization methods that are based on the atomic structure, this new assembling algorithm uses color coded 3D molecular surfaces of the nucleotides as the building blocks for DNA assembling. This new approach, not only reduces the amount of graphical objects and, consequently, makes the rendering faster, but also makes it easier to visually identify the nucleotides in the DNA strands. The algorithm used to triangulate the molecular surfaces of the nucleotides building blocks is also a novelty presented as part of this dissertation. This new triangulation algorithm for Gaussian molecular surfaces introduces a new mechanism that divides the atomic structure of molecules into boxes and spheres. This new space division method is faster because it confines the local calculation of the molecular surface to a specific region of influence of the atomic structure, not taking into account atoms that do not influence the triangulation of the molecular surface in that region. This new method also guarantees the continuity of the molecular surface. Having in mind that the aim of this dissertation is to present a complete set of methods for plasmid DNA visualization and simulation, it is also proposed a new deformation algorithm to be used for plasmid DNA Monte Carlo simulations. This new deformation algorithm uses a 3D polyline to represent the plasmid DNA conformation and performs small deformations on that polyline, keeping the segments length and connectivity. Experiments have been performed in order to compare this new deformation method with deformation methods traditionally used by Monte Carlo plasmid DNA simulations These experiments shown that the new method is more efficient in the sense that its trial acceptance ratio is higher and it converges sooner and faster to the elastic energy equilibrium state of the plasmid DNA molecule. In sum, this dissertation successfully presents an end-to-end set of models and algorithms for plasmid DNA geometric modelling, visualization and simulation. | pt_PT |
| dc.identifier.tid | 101477490 | |
| dc.identifier.uri | http://hdl.handle.net/10400.6/4039 | |
| dc.language.iso | eng | pt_PT |
| dc.subject | Modelação geométrica de ADN | pt_PT |
| dc.subject | Simulação de ADN plasmídeo | pt_PT |
| dc.title | Geometric modeling, simulation, and visualization methods for plasmid DNA molecules | pt_PT |
| dc.type | doctoral thesis | |
| dspace.entity.type | Publication | |
| person.familyName | Raposo | |
| person.givenName | Adriano | |
| person.identifier.ciencia-id | F910-7A7C-E6DD | |
| person.identifier.orcid | 0000-0002-7432-7474 | |
| rcaap.rights | openAccess | pt_PT |
| rcaap.type | doctoralThesis | pt_PT |
| relation.isAuthorOfPublication | 954acde5-2baa-40af-b3bd-da49f0f28f60 | |
| relation.isAuthorOfPublication.latestForDiscovery | 954acde5-2baa-40af-b3bd-da49f0f28f60 | |
| thesis.degree.name | Doutoramento em Engenharia Informática | pt_PT |