University of Ioannina
Ανάπτυξη ευφυών μεθόδων για την ανάλυση και διαχείριση γενετικής πληροφορίας
Abstract
dc:descriptionIn this PhD thesis a methodology for protein fold recognition is presented, which leads to the recognition of protein structure. The method is fully automated and given an aminoacid sequence of unknown structure it can classify it into one of the known protein fold categories with high success rate. The methodology has very low computational cost and good results compared to the existing methods which are mentioned in bibliography.More specifically, in the first chapter of the thesis a brief introduction about the genetic material and the proteins is presented. The information flow in molecular biology is described and it is also explained why the knowledge of proteins’ structure is so significant for the understanding of the function of human’s organism. Moreover, the contribution of computational methods to the analysis of genetic information and to the extraction of useful biological knowledge is also described. In the second chapter the difficulties that appear in fold recognition process are initially presented. Then follows a review of the methods for fold recognition that appear in the literature and the detection of their weak points.In the third chapter the proposed methodology is introduced. The Hidden Markov Model used for protein fold recognition is described and is explained how it is trained with protein sequence data. Then the classification process for the proteins of unknown structure is analysed. Finally, the classification competence of the model is evaluated with the use of sequences coming from protein databases, along with the comparison of the extracted results with the those of other methods that have been proposed in literature.In the fourth chapter specific improvements are introduced to the employed model. These improvements are presented in two stages. In the first stage the structure of the model changes and in the second stage the way of scoring the protein sequences against the model changes. These changes lead simultaneously to simplification of the model and to better ability of classifying proteins of unknown structure. In each stage the efficiency of the new form of the model is evaluated with the use of protein sequence data. The extracted results are compared to those calculated when the initial form of the model is applied in the same data.In the fifth chapter a hybrid system is presented which is an alternative approach for classification of proteins of unknown structure. Specifically, a combination of use of a Markov chain and the hidden Markov model takes place along with an optimization method which yields even better results.In the sixth chapter the conclusions of the PhD thesis are presented. More specifically, the advantages and the disadvantages of each stage of the proposed methodology are analysed. Moreover, the perspectives of future work for extending and improving the system are investigated.
Degree
thesis:*- Grantor dc:publisher
- University of Ioannina
- Year dc:date
- 2012
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Lampros, Christos
- Λάμπρος, Χρήστος
Subjects
dc:subject × 20- Πρωτεΐνες
- Κρυμμένο μαρκοβιανό μοντέλο
- Αναγνώριση πρωτεϊνικής αναδίπλωσης
- Αμινοξέα
- Τάξεις
- Συγκριτική μοντελοποίηση
- Μαρκοβιανή αλυσίδα
- Πρόβλεψη δομής
- Proteins
- Hidden Markov model
- Protein Fold Recognition
- Aminoacids
- Classes
- Comparative modelling
- Markov chain
- Structure prediction
- Φυσικές Επιστήμες
- Βιολογία
- Natural Sciences
- Biological Sciences
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/28465