Dissertation/ Thesis
Evolutionary Algorithms for Multiple Sequence Alignments
| Τίτλος: | Evolutionary Algorithms for Multiple Sequence Alignments |
|---|---|
| Συγγραφείς: | Naznin, Farhana |
| Στοιχεία εκδότη: | UNSW, Sydney |
| Έτος έκδοσης: | 2011 |
| Συλλογή: | UNSW Sydney (The University of New South Wales): UNSWorks |
| Θεματικοί όροι: | Amino Acid Sequences, Deoxyribonucleic Acid, Ribonucleic Acid, Dynamic Algorithm / Programming, Pairwise Sequence Alignment, Guide Tree / Progressive Alignment Approach, Multiple Sequence Alignment, Iterative Approach, Genetic Algorithm |
| Περιγραφή: | Multiple sequence alignment (MSA) is a valuable technique for studying molecular evolution, analysing sequence structure, and drug design. To date, in the literature, many algorithms for solving the MSA problems, based on progressive, iterative and stochastic approaches, have been proposed. As their performances vary depending on the characteristics of the test problems, there is a need for an effective algorithm that consistently performs better over a range of problems. Therefore, the objective of this thesis was to develop a new algorithm for solving MSA problems more effectively. However, our systematic analyses of the existing approaches encouraged us to propose not only a new algorithm but also the sequential development of increasingly effective algorithms. Firstly, this thesis presents a new iterative method, namely MSA-IT, which has two new mechanisms: the first generates guide trees with randomly selected sequences; and the second shuffles the sequences inside these trees, within an iterative process, to generate new and better guide trees. This method is simple and fast. Its performance was tested on standard benchmark problems and, compared with well-known algorithms, was convincing. However, as its solutions made clear that there was room for further improvement, we developed another algorithm based on a genetic algorithm with a progressive alignment approach, namely GAPAM. In GAPAM, the initial population is generated using the two mechanisms used in MSA-IT, and two new genetic search operators are implemented. To test its performance on benchmark datasets, the results from GAPAM were compared with those from MSA-IT as well as the well-known algorithms. The comparisons showed that GAPAM not only provides better solutions than MSA-IT and the other algorithms for most of the test problems, but also achieves an overall better performance. However, analyses of the results uncovered the fact that, for sequences of longer length, GAPAM was trapped in local optima. To overcome this local optima problem, ... |
| Τύπος εγγράφου: | doctoral or postdoctoral thesis |
| Περιγραφή αρχείου: | application/pdf |
| Γλώσσα: | English |
| Relation: | https://hdl.handle.net/1959.4/51535; https://doi.org/10.26190/unsworks/15176 |
| DOI: | 10.26190/unsworks/15176 |
| Διαθεσιμότητα: | https://hdl.handle.net/1959.4/51535 https://unsworks.unsw.edu.au/bitstreams/8e11d1c2-0ff5-4ff1-938a-d0b70ddb1436/download https://doi.org/10.26190/unsworks/15176 |
| Rights: | open access ; https://purl.org/coar/access_right/c_abf2 ; CC BY-NC-ND 3.0 ; https://creativecommons.org/licenses/by-nc-nd/3.0/au/ ; free_to_read |
| Αριθμός Καταχώρησης: | edsbas.246EB642 |
| Βάση Δεδομένων: | BASE |
| DOI: | 10.26190/unsworks/15176 |
|---|