| Summary: | 博士 === 國立清華大學 === 資訊工程學系 === 95 === The evolutionary mystery of species has been extensively investigated within the scientific
area and there is a great quantity of approaches towards attacking this issue. Here we
exploit the combinatorial method for tracing evolutionary paths based on the model of
genome rearrangements, which provides appropriate explanations for the disruption of
gene order in two or more genomes by using large-scale mutations. Since these mutations
are very rare and easy to detect, the study in genome rearrangements often allows us to
conduct plausible evolutionary scenarios composed of the considered mutations step by
step.
At first, we try to depict the evolutionary story using two global mutations of reversals
and block-interchanges. A reversal event reverses the order of the elements in
a segment and flips their orientations, and a block-interchange event swaps two nonoverlapping
segments. We deal with two problems: One is to find the minimum number
of block-interchanges for transforming one permutation into another. By using the algebraic
approach, we can obtain an optimal solution in O(n) time, where is the number
of block-interchanges used. The other is a novel problem considering the reversals and
block-interchanges together and seeking a scenario with the minimum weight-sum. Based
on the breakpoint graph approach, we propose an algorithms running in O(n^2) time to
exactly solve it. Our obtained solution has not only the minimum weighted sum but also
the minimum number of block-interchanges used among all optimal solutions, which agrees
with the biological observation. Moreover, such a minimum number/weight-sum is often
regarded as the evolutionary distance between two species.
The second part of this thesis is devoted to close the gap between the theoretical
analysis and biological application via our two websites allowing DNA sequences as its
inputs and computing the distance for each pair of input sequences. Besides distances,
our websites give a possible scenario between any two sequences and a corresponding
phylogenetic tree. Therefore, biologists can investigate the evolutionary relationship for a
group of species in a more convenient and systemic manner. Subsequently, to validate our
websites, we have tested them with two sets of sequences, human vibrio pathogens and
-
proteobacteria, for detecting their evolutionary relationships. Our experimental result on
vibrio pathogens coincides with the previous one using a different comparative genomics
approach, and the other on
-proteobacteria receives a better outcome than before.
Furthermore, we also contribute to two book chapters. One arises in the attempt
to be a tutorial-type approach for our website including the tips, tricks and solutions to
do the experiments. The other is a survey article on the introduction of interesting and
developing topics in genome rearrangements, which enables researchers to grasp typical
analytic methods and have a global view in genome rearrangements. In addition, we give
a clear table to show the current progress in genome rearrangements, and from the table
one can see not only the hot research topics and possible research directions, but also
that genome rearrangements is still one of the most promising research in computational
molecular biology.
|
|---|
