Modeling and simulation of protein folding and proteostasis mechanisms
Date Issued
January 16, 2023
Type
Διδακτορική Διατριβή
Abstract
The problem of protein folding is one of the many unsolved paradigms in biochemistry and has been considered as one of the most important challenges to solve, particularly because the complicated protein structure makes it challenging to predict the optimal folding patterns. Protein folding has been projected as a fundamental issue in the research of molecular sciences as a result of the identification of several diseases as protein folding disorders, the explosion of information regarding the genome, and the requirement for effective methods to predict the structure of proteins. The study of probable variations in the tertiary structure of proteins may give new insights linked to the relationship between mutations and protein functions. This is because mutations have the potential to have an effect on the functionality of proteins. We discovered through the use of protein simulations that proteins are extremely dynamic and, in the absence of a ligand, sample a wide variety of conformational states, including both open and closed configurations. These states are selectively stabilized by the binding of agonists and antagonists, respectively.
This PhD thesis describes a structural-bioinformatics method as a means of identifying candidate proteins and particular mutations that may be associated with the development and progression of disease. The primary structures and the mutations that are associated with them were retrieved from public databases so that an effect analysis could be performed. Effective approaches for structure prediction were used to make predictions about each structure. A comparison of the structural prediction tools' performance was successfully carried out. Classic bioinformatics methods (TM-Score, RMSD, etc.) and established shape-based descriptors acquired from object recognition methodologies were both utilized in the comparative analysis of altered and unmutated protein sequences. The structures were analyzed using unsupervised methods in order to discover groups of mutations having a similar impact on the structure. The method that was applied was highly efficient, and it was able to successfully introduce a new facet into the process of determining pathogenicity.
In this doctoral thesis, 110 mutant structures of PSEN1 were characterized as being pathogenic. Additionally, six unclear structures were characterized for APP, nine for PSEN2, and four for MAPT protein.
Subjects