Use this URL to cite or link to this record in EThOS:
Title: Prediction and analysis of nucleosome positioning in genomic sequences
Author: Hasan, S.
Awarding Body: University of Cambridge
Current Institution: University of Cambridge
Date of Award: 2003
Availability of Full Text:
Full text unavailable from EThOS.
Please contact the current institution’s library for further details.
One theory suggests that DNA sequences, which are intrinsically “curved”, can position nucleosomes. In a previous study, using “cyclical” hidden-markov models, it had been suggested that a 10 periodic occurrence of the [VWG] motif could have such an effect and could help nucleosomes to be positioned in human exons. This work was extended in this thesis. 60% of human genomic sequences were seen to be covered in apparently weak 9-10 bp periodic patches of [CWG]. [CWG]-dense regions were seen to alternate with regions which were rich in [W] motifs in human. However, the pattern was not the same in mouse. Another theory suggests that highly flexible or highly rigid DNA sequences may favour or disfavour nucleosome formation respectively. The locations of such patterns were investigated in human sequences using the wavelet technique. This approach identified confined periodic patterns (in the range of 80-200 bp) of rigidity in human genomic sequences; the patterns themselves were, however, mainly consequences of alu repeat-clustering. However, the same analysis in the mouse genome indicated that such a mechanism for positioning nucleosomes was not conserved and therefore unlikely. A different approach to model nucleosomes was to train weighted DNA matrices from experimentally-mapped nucleosomes datasets. This technique gave some encouraging results (one model showing 100% accuracy at 40% coverage), but was restricted by the limited size of the datasets. Overall the conclusion is that there is some evidence for sequence specific nucleosome positioning, but that more experimental data is needed to build and evaluate practical and predictive computational models.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available