Abstract:
Fixed-parameter algorithms offer a constructive and powerful approach
to efficiently obtain solutions for NP-hard problems combining two
important goals: Fixed-parameter algorithms compute optimal solutions
within provable time bounds despite the (almost inevitable)
computational intractability of NP-hard problems. The essential idea
is to identify one or more aspects of the input to a problem as the
parameters, and to confine the combinatorial explosion of
computational difficulty to a function of the parameters such that the
costs are polynomial in the non-parameterized part of the input. This
makes especially sense for parameters which have small values in
applications. Fixed-parameter algorithms have become an established
algorithmic tool in a variety of application areas, among them
computational biology where small values for problem parameters are
often observed. A number of design techniques for fixed-parameter
algorithms have been proposed and bounded search trees are one of
them. In computational biology, however, examples of bounded search
tree algorithms have been, so far, rare.
This thesis investigates the use of bounded search tree algorithms for
consensus problems in the analysis of DNA and RNA data. More
precisely, we investigate consensus problems in the contexts of
sequence analysis, of quartet methods for phylogenetic reconstruction,
of gene order analysis, and of RNA secondary structure comparison. In
all cases, we present new efficient algorithms that incorporate the
bounded search tree paradigm in novel ways. On our way, we also obtain
results of parameterized hardness, showing that the respective
problems are unlikely to allow for a fixed-parameter algorithm, and we
introduce integer linear programs (ILP's) as a tool for classifying
problems as fixed-parameter tractable, i.e., as having fixed-parameter
algorithms. Most of our algorithms were implemented and tested on
practical data.