Applications and Extensions of Network Alignment
|has title::Applications and Extensions of Network Alignment|
|Student name:||student name::Marlies van der Wees|
|Second supervisor:||Gunnar Klau|
|Second reader:||has second reader::Jaap Heringa|
Recently enormous amounts of data involving interactions in humans and model organisms have become available [Sharan et al., 2006]. Unlike sequence data, which can be represented as a string over some finite alphabet, interactions are captured in a network consisting of nodes connected by edges. The nodes in a network represent the entities under consideration and the edges represent interactions between the nodes. Examples of networks include protein-protein interaction (PPI), gene-regulatory, and signal transduction networks. In order to elucidate functionally significant interactions, it is imperative to have a way to identify similarity among networks. Network alignment allows one to accomplish this, and thereby, similarly to sequence alignment, to identify evolutionary conservation and functional significance by means of comparative analysis.
In this project we want to apply and extend ideas from [El-Kebir et al., 2011] in order to facilitate application of the method to protein-protein interaction networks from the STRING database [Szklarczyk et al., 2011]. To this end, we plan to develop an easily-accessible webserver. This will be done in collaboration with ACTA. Together with the Medical Bioinformatics and e-Bioscience (AMC) we will apply and extend the method to study the transferability of clinical experimental results from model organisms to human. Challenges include finding both similar and dissimilar regions between networks of co-expressed genes. For that purpose we will consider a novel scoring function. In addition, we will need to extend the method to deal with the large scale of the data.