'Computational grids' are spatially distributed heterogeneous collections of computing resources. For example, Ontario's SHARCNET is a powerful networked system of compute clusters spread out over Ontario's Universities.

If the computing power available in such grids can be unleashed effectively, large scientific problems can be solved that would be unsolvable using stand-alone machines.

In particular, many bioinformatics problems pose a dire computational challenge, both because the amount of data to be processed can be huge, and because the processing algorithms can be extremely complex.

In this project you will develop techniques that allow to apply the power of computational grids to complex parallel bioinformatics problems.

You will build upon the success of the TaskSpaces software framework for scientific computing on computational grids, and initial application to an RNA folding statistics problem, see also 'A Software Framework for Parallel Bioinformatics on Computational Grids' [pdf].

Several directions are possible for this project, and may include further development of a generally usable grid computing software framework for parallel bioinformatics, for instance allowing the use of problem specification in the Python language, which is becoming the lingua franca of bioinformatics. Scalability, fault-tolerance, and security are other important aspects for a grid computing framework that may be further investigated. Alternatively, you may prefer to work closer to bioinformatics applications, and may as such contribute to solving bioinformatics problems that require large computational resources. Some applications involve extensive use of combinatorics, search algorithms, and other applied mathematics techniques, which may be other interesting topics to pursue.

Computing resources provided by SHARCNET will be studied and employed extensively in this project.

Software development and bioinformatics applications will involve collaboration with Prof. Rob Knight of the MCDB Department of the University of Colorado at Boulder, USA. For the software environment development aspects of the project, a good knowledge of Java and/or Python and a strong interest in (parallel and grid) computing are required. The emphasis of your project can either be on the software framework development, or on parallel bioinformatics applications and the mathematics required to solve them, or both.

This topic is suitable for a PhD or Master's thesis. Please contact Hans De Sterck if you are interested.

Created by Hans De Sterck.
Department of Applied Mathematics, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada.
Phone: 1-519-888-4567 ext 7550, Fax: 1-519-746-4319, E-mail: hdesterck@math.uwaterloo.ca.
Office: MC 5016. campus map