Back to search

University of Patras

Υπολογιστικές μέθοδοι για σχεδιασμό και ανάπτυξη φαρμάκων πολλαπλών στόχων για σύνθετες ασθένειες

Abstract

dc:description

Drug design has largely followed a reductionist’s approach, which entails finding a compound that reacts with a single molecular target, as well as a phenotypic based discovery process which does not focus on the drug’s mode of action. Although this approach has served the industry in the past and reached its zenith in the 1990’s, it has been producing diminishing returns since the turn of the century. In the last years, there has been a paradigm shift towards multi-target based discovery, aiming to address the productivity gap in drug development, by focusing on the elucidation of the disease’s causes on a molecular level, as well as the exact mode of action of a drug, while leveraging advancements in high-throughput technologies. The picture here is extremely complex. Molecular interactions form complex networks that regulate biological functions, resulting in specific phenotypes. These networks are generally robust; however, a variety of environmental and genetic factors may produce perturbations which can result in the rise of a disease. Disease networks are also robust. Drugs also operate on a molecular level, resulting in perturbations propagated along different levels of organization to cure a disease or alleviate its symptoms. Systems pharmacology brings the promise of revolutionizing the drug discovery process, while at the same time catalysing the translation of pharmacogenomics applications to the clinical environment, which has been lagging behind despite the recent wave of groundbreaking research on genomics implications in disease. The aim of the doctoral thesis was to design and develop novel computational methods, open source software and visualizations that can be used along the main application areas of systems pharmacology: characterization of drug mechanism of action, identification of novel drug targets, drug repurposing and side-effect prediction. The first chapter presents an in-depth review of computational methods that are utilized in each of the application areas of systems pharmacology. The second chapter presents a method for the identification of time-varying differentially expressed subpathways. The method offers the explanatory power to offer focused temporal snapshots of the mechanisms perturbed under specific conditions, offering a holistic view of the dynamics of molecules participating in each pathway, by considering the time aspect of the activation of a subpathway, the type of regulation between subpathway members and ultimately the topological aspects of the activated subpathways, which portray the capacity of each subpathway to affect systemic changes within multiple pathways. The third chapter presents a method for the identification of differentially expressed subpathways within multi-pathway networks. By defining the subpathway within a combined multi-pathway network, the method takes into account pathway crosstalks, a field of research which has gained significant attention. In order to extract differentially expressed subpathways, the method utilizes an array of network analysis methods. Differential expression analysis on the gene level is possible via an array of well-known differential expression analysis methods. Finally, each subpathway is associated with functional and pharmacological terms that characterize, regulate or are connected with its expression. The fourth chapter presents a method for the identification of differentially expressed subpathways using a consensus approach. The method refines a combined multi-pathway network, by utilizing differentially expressed genes that exhibit coordinated expression level with their neighbors, and topologically significant non differentially expressed genes that accommodate the propagation of the biological signal, in order to reduce the density of the combined network in a biologically meaningful manner. Finally, the refined network is decomposed to a list of differentially expressed subpathways using an array of network analysis methods and checked for enrichment in functional terms such as drugs, diseases and side-effects. The fifth chapter presents a method for the identification of differentially expressed subpathways using a bilevel consensus scoring of pathway topology and gene expression in single pathway networks, by combining differential expression evidence and ranking the connectivity of each subpathway within a multi-pathway network, in order to produce candidate biomarkers for pharmacological intervention, ranked according to their potential to induce side-effects. The sixth chapter presents a novel visualization theme for multi-level subpathways that are constructed by subpathway analysis methods. The theme is interactive and merges a network structure with the information-rich view of circular plots. It depicts a multi-layer network structure with functional associations with terms of interest and topological associations captured by network centrality measures. It is accessible via a graphical environment that can extend the explanatory capabilities of subpathway analysis methods implemented as R-packages.

Degree

thesis:*
Grantor dc:publisher
University of Patras
Year dc:date
2020

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Balomenos, Panagiotis
  • Μπαλωμένος, Παναγιώτης

Subjects

dc:subject × 12

Rights

dc:rights
Statement dc:rights
  • BY_NC_ND
Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/49213

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Balomenos, Panagiotis; Μπαλωμένος, Παναγιώτης. Υπολογιστικές μέθοδοι για σχεδιασμό και ανάπτυξη φαρμάκων πολλαπλών στόχων για σύνθετες ασθένειες. University of Patras, 2020. http://hdl.handle.net/10442/hedi/49213