|
Identifier |
000355414 |
Title |
Υπολογιστικά εργαλεία για τον εντοπισμό και μελέτη της λειτουργίας των αλληλεπιδράσεων ρυθμιστικών δικτύων |
Alternative Title |
Development of computational tools for the identification and analysis of functional interactions of regulatory networks |
Author
|
Μανιουδάκη, Μαρία E.
|
Thesis advisor
|
Τσιώτης, Γεώργιος
Ποϊράζη, Παναγιώτα
|
Abstract |
Yeast cells live in a constantly changing environment that requires the continuous adaptation of their genomic program in order for them to survive, proliferate and sustain their homeostasis. The regulation of genes can be represented as a complex network of interactions that consists of the proteins (transcription factors) and the genes that these proteins regulate as a response to specific signals. The development of high-throughput technologies has contributed to the accumulation of a large amount of data that can be used to build intracellular networks using a holistic approach. This processing requires powerful computational methods able to combine different types of data and identify quantitative and qualitative interactions among their components.
Aim of the present work is to use the appropriate computational methods, specifically module identification methods and Artificial Neural Networks to combine available high-throughput experimental data in order to derive the structure of regulatory networks, identify quantitative interactions among their components and provide evidence for the mechanism of their regulation. The method was initially applied to experimental data regarding the response of Saccharomyces cerevisiae cells in various environmental changes but it is believed that it can be applied to genomic data of any organism, or to various cellular responses of the same organism.
In brief, this work shows that: (a) the expression profile of at least 17/91 Saccharomyces cerevisiae genes related to stress can be predicted, given the expression profile of transcription factors that they exert their regulatory activity two layers upstream in the regulatory cascade in which these genes participate, (b) there is a delayed response between the expression of the second layer transcription factors and the target genes in 30/38 networks, (c) three networks that include known protein interactions can be successfully modeled as well as (d) the expression of two genes in perturbed networks and (e) is possible to identify six novel interactions that were not previously related to stress.
The outcome of this work shows that a combination of gene expression data, protein-DNA interaction data and protein-protein interaction data analyzed by Artificial Neural Networks can infer and model biological networks, correlate quantitative interactions among regulators and regulatory genes, identify time-delayed regulatory interactions and suggest a probable mechanism of gene expression control under multiple conditions
|
Physical description |
178 σ. : έγχ. εικ. ; 30 εκ. |
Language |
Greek |
Subject |
Artificial neural networks |
|
Environmental stress |
|
Functional modules |
|
Gene expression analysis |
|
Protein-DNA interactions |
|
Regulatory networks |
|
Saccharomyces Cerevisiae |
|
Time delays |
|
Αλληλεπιδράσεις πρωτεϊνης-DNA |
|
Ανάλυση γονιδιακής έκφρασης |
|
Περιβαλλοντικό στρες |
|
Ρυθμιστικά δίκτυα |
|
Ρυθμιστικές υπομονάδες |
|
Τεχνητά νευρωνικά δίκτυα |
|
Χρονικές καθυστερήσεις |
Issue date |
2010-01-22 |
Collection
|
School/Department--School of Sciences and Engineering--Department of Chemistry--Doctoral theses
|
|
Type of Work--Doctoral theses
|
Views |
360 |