Quantitative analysis of Drosophila melanogaster gap gene expression data reveals valuable information about the nature and strengths of interactions in the gap gene network. We first explore different models for fitting the spatiotemporal gene expression data of Drosophila gap gene system and validate our results by computational analysis and comparison with the existing literature. A fundamental problem in systems biology is to associate these results with the inherent cause of gene regulation, namely the binding of the transcription factors (TF) to their respective binding sites. In order to relate these expression-based estimates of gap gene regulation with the sequence-based information of TF binding site composition, we also explore two related problems of (i) finding a set of regulatory weights that is proportional to the binding site occupancy matrix of the transcription factors in current literature and (ii) finding a set of position weight matrices of the TFs that produce a new binding site occupancy matrix showing a greater level of proportionality with our regulatory weights. Our solution to the first problem yielded a regulatory weight matrix incapable of explaining the true causes of gene expression profile despite its relative numerical accuracy in predicting the gene expressions. On the other hand, the second optimization problem could be solved up to a reasonable level of accuracy, but further analysis on the result demonstrated that this optimization problem may be under-constrained. We devise a simple regularization strategy that helps us to reduce the under-constrained nature of the problem.
| Identifer | oai:union.ndltd.org:LACETR/oai:collectionscanada.gc.ca:QMM.112321 |
| Date | January 2007 |
| Creators | Al Zamal, Faiyaz. |
| Publisher | McGill University |
| Source Sets | Library and Archives Canada ETDs Repository / Centre d'archives des thèses électroniques de Bibliothèque et Archives Canada |
| Language | English |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Format | application/pdf |
| Coverage | Master of Science (School of Computer Science.) |
| Rights | All items in eScholarship@McGill are protected by copyright with all rights reserved unless otherwise indicated. |
| Relation | alephsysno: 002710949, proquestno: AAIMR51358, Theses scanned by UMI/ProQuest. |
Page generated in 0.0025 seconds