National and Kapodistrian University of Athens
Ανίχνευση νέων φαρμακολογικών στόχων μέσω υπολογιστικής μελέτης και ανάλυσης της τρισδιάστατης δομής των πρωτεϊνών
Abstract
dc:descriptionDrugs are a valuable commodity used for the prevention and treatment of diseases that threaten humans during their lifetime. Since their rise, struggle for survival has also included protection from hardships and diseases apart from food retrieval. In ancient Greece, Hippocrates used to teach, "Food should be your medicine and your medicine should be your food''. A great reduction in novel pharmacological compounds is observed in recent years, despite the great efforts that take place for the treatment of diseases and pathological conditions. This problem can be pinpointed to the lack of knowledge for novel drug targets. Drug targets can include host proteins, mainly enzymes and receptors, along with antibodies. In other cases, though, drug targets can be viral or parasite proteins that infect humans. Proteins carry one or many specific functional center for endogenous ligand binding or interaction with other proteins. Drug targeting involves mainly the functional center, although it is possible to observe binding in other protein sites. The identification of novel drug targets is heavily associated with the study of the molecular target's 3D structure. Nowadays, however, technological evolution in bioinformatics allows the use of 3D structure of proteins or protein-ligand and protein-protein complexes in order to identify new drug targets. In the current Ph.D. study a specific pipeline was developed in order to identify novel drug targets. This pipeline includes a computational study and analysis of the protein 3D structure. Protein amino-acid sequences are characteristic of their genetic diversity, while their 3D structure and molecular architecture determines their function, and in the 3D protein complexes the type of interactions they exhibit. These three different types of information are important for a better understanding of structural and functional protein characteristics and for finding new pharmacological targets. Many attempts have been made until now to associate information from amino-acid sequences, 3D structures, and 3D protein complexes in order to elucidate the unknown mechanisms underlying their function. In the current Ph.D. study, four different biological datasets of different types of proteins are used to design a complete pipeline in order to find new drug targets. The four protein datasets analyzed during the current study are antibodies, nuclear receptors, viral enzymes of the Flaviviridae family of viruses, and RNA polymerase II of the parasite that causes malaria. The results produced are significant and novel. Four new methodologies have been developed in order to study and analyze the biological information proteins store in their structure and amino-acid sequence used during the process of the computational drug target elucidation pipeline. The first methodology describes a new way of identifying novel drug targets through the use of conserved patterns signaling patterns in amino-acid sequences. The second methodology describes a new way of protein clustering and association through the use of representative physiochemical models and unsupervised learning techniques, including the k-means algorithm and neuronal networks towards estimating signaling patterns extension. The third methodology was developed for analyzing the 3D structure and describes a new way of associating and clustering 3D structures by using hybrid distance tables and ranking clustering techniques such as neighbor-joining and UPGMA. Last but not least, during the study and analysis of the 3D structure of protein complexes’, a new method of digitally coding protein interaction into interaction patterns / signaling patterns was developed and used for the clustering of studied proteins into groups based on antibody-interaction patterns. A new application has been developed during the aforementioned pipeline called “antisoma” that focuses on the protein information regarding antibodies. This application is a fully functional platform containing the variable antibody regions with important information regarding their containing proteins. The user has the option to extract different datasets from amino-acid sequences, apply different filters based on the antibody numbering system, and estimate more than 300 physiochemical attributes as a signaling pattern for the protein fractions that are included in the variable regions. The conclusions drawn from the various stages of the computational analyses are given at the end of each chapter. It is also important to state that using the methodologies and the application developed, the research team has managed to cluster the biological information of several protein datasets into distinct groups . In conclusion, the suggested computational study has identified an important number of new pharmacological targets by applications on distinct protein datasets.
Degree
thesis:*- Grantor dc:publisher
- National and Kapodistrian University of Athens
- Year dc:date
- 2021
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Papageorgiou, Louis
- Παπαγεωργίου, Λούης
Subjects
dc:subject × 28- Λούης
- Παπαγεωρίου
- Υπολογιστική βιολογία
- Δομική βιοπληροφορική
- Βιοπληροφορική
- Πρότυπα σήματος
- Φαρμακολογικοί στόχοι
- Σχεδιασμός αλγορίθμων
- Louis
- Papageorgiou
- Computational biology
- Structural bioinformatics
- Pipeline design
- Data mining
- Algorithms design
- Signaling patters
- Bioinformatics
- Pharmacological targets
- Επιστήμη Ηλεκτρονικών Υπολογιστών και Πληροφορική
- Φυσικές Επιστήμες
- Άλλες Επιστήμες Μηχανικού και Τεχνολογίες
- Επιστήμες Μηχανικού και Τεχνολογία
- Βιοτεχνολογία και Εφαρμοσμένη μικροβιολογία
- Computer and Information Sciences
- Natural Sciences
- Other Engineering and Technologies
- Engineering and Technology
- Biotechnology and Applied Microbiology
Rights
dc:rights- Statement dc:rights
-
- BY_4_0
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/49567