Use this URL to cite or link to this record in EThOS:
Title: Statistical Analysis of cDNA Microarray Directed by Gene Function
Author: Soneji, Sharnit
ISNI:       0000 0001 3470 6974
Awarding Body: Birkbeck (University of London)
Current Institution: Birkbeck (University of London)
Date of Award: 2007
Availability of Full Text:
Access from EThOS:
Microarrays allow the expression level of thousands of genes to be measured simultaneously. This study will address analytical issues predominantly concerned with eDNA arrays. These include normalisation and data preprocessing, leading to an assessment of cluster analysis and the integration of database information to elucidate functional classes of biological relevance. This is then extended further to classify genes of unknown function using Markov Random Fields. I Modelling of uneven surface trends were considered in a new 2D-normalisation method which outperformed the popular loess method which concentrated on printing pin effects. With respect to cluster analysis, a new method to identify the number of clusters in higher dimension da~a is proposed which provides a visual way of determining at which point over-fitting of the data will occur. Once partitioned, functional information was incorporated to find enrichment of classes in clusters using a new application of X2 bootstrapping, which provides a very robust way of identifying these groups. A novel use of correspondence analysis was applied to the contingency tables produced from the cluster over class analysis which was used to show that functionally related groups were acting in concert when scrutinising the projection' of these classes onto three dimensions. The last part of this study attempted the use of Markov Random Fields to assign function to genes of unknown function using M. tuberculosis and E. coli data. The ability to determine function from the data used in this work was limited, but the method implemented in this study showed improvement over previous attempts.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available