Computational optimization of peptides using
evolutionary algorithms
Evolutionary algorithms , for instance 'genetic algorithms (GA)', evolve based on a starting population better solutions for a certain optimization problem.
In our case the starting population consist of peptides, differing in their three-dimensional structure (folds) and sequence. Applying evolutionary optimizations (more precisely multi-objective optimizations) we try to select the best candidate peptides, which form a stable fold and fulfill certain tasks, like acting as a drug.
We have to tackle a two-level optimization problem: (1) we have to find an optimal evolutionary algorithm, and (2) to apply this algorithm to the optimization of peptides.
For the first step, which is very compute-intensive, we use simplified peptide models that conserve peptide features and require less computation time. Applying these simplified models we can perform an extensive search/screen for evolutionary parameters that may be tested on the more detailed peptide models in the second step.
The resulting peptide candidates may be synthesized and serve as e.g. drug targets.
1st supervisor: Prof. Dr. Daniel Hoffmann
2nd supervisor:
Institute:
Department of Structural and Medicinal Biochemistry, Centre for Medical Biotechnology (ZMB), University of Duisburg-Essen
Telephone: +49 (0)201 - 183 - 3964
E-mail: reda.rawi@uni-due.de |
