Services for biological network feature detection

Gollapudi, Venkata Lakshmi Sirisha (2010) Services for biological network feature detection. PhD thesis, University of Nottingham.

PDF - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Download (4MB) | Preview


The complex environment of a living cell contains many molecules interacting in a variety of ways. Examples include the physical interaction between two proteins, or the biochemical interaction between an enzyme and its substrate. A challenge of systems biology is to understand the network of interactions between biological molecules, derived experimentally or computationally. Sophisticated dynamic modelling approaches provide detailed knowledge about single processes or individual pathways. However such methods are far less tractable for holistic cellular models, which are instead represented at the level of network topology.

Current network analysis packages tend to be standalone desktop tools which rely on local resources and whose operations are not easily integrated with other software and databases. A key contribution of this thesis is an extensible toolkit of biological network construction and analysis operations, developed as web services. Web services are a distributed technology that enable machine-to-machine interaction over a network, and promote interoperability by allowing tools deployed on heterogeneous systems to interface. A conceptual framework has been created, which is realised practically through the proposal of a common graph format to standardise network data, and the investigation of open-source deployment technologies. Workflows are a graph of web services, allowing analyses to be carried out as part of a bigger software pipeline. They may be constructed using web services within the toolkit together with those from other providers, and can be saved, shared and reused, allowing biologists to construct their own complex queries over various tools and datasets, or execute pre-constructed workflows designed by expert bioinformaticians.

Biologically relevant results have been produced as a result of this approach. One very interesting hypothesis has been generated regarding the regulation of yeast glycolysis by a protein found to interact with seven glycolytic enzymes. This has implied a potentially novel regulatory mechanism whereby the protein in question binds these enzymes to form an 'energy production unit'. Also of interest are workflows which identify termini (system inputs and outputs), and cycles, which are crucial for acquiring a physiological perspective on network behaviour.

Item Type: Thesis (University of Nottingham only) (PhD)
Supervisors: Hodgman, T.C.
Greenhalgh, C.M.
Subjects: Q Science > QH Natural history. Biology > QH301 Biology (General)
Faculties/Schools: UK Campuses > Faculty of Science > School of Biosciences
Item ID: 13022
Depositing User: EP, Services
Date Deposited: 14 Jan 2013 09:08
Last Modified: 17 Dec 2017 17:02

Actions (Archive Staff Only)

Edit View Edit View