The main purpose of our contribution is to transfer the knowledge from local solutions into production utilization using available grid research area built around the EGEE infrastructure, and thus permitting scaling up the performed computations. The faster exploration of conformational space achieved by the Grid utilization forms an important step towards answering more challenging research questions. Moreover, our implementation serves as a show case of high-throughput type of computational chemistry calculations running within the worldwide distributed infrastructure, thus demonstrating its suitability for such kind of computations.
Conclusions and Future Work
Here we present an advanced way how to perform efficient exploration of conformational space of biomolecules within the worldwide EGEE/EGI environment. Our extension to the approach, already tested using single CPU and MPI-parallelized implementations, allows routine utilization within the worldwide Grid, thus boosting the research in the field of flexibility and folding of biomolecular systems.
The program grid-CICADA is composed of two main components: a) control unit and b) explorers. The control unit collects information about alreadythe searched conformational space. It optimizes found interconversion pathways among conformers and rationalizes the strategy of further space exploration. It also manages explorers, which perform a simple search in a limited conformational subspace. Since individual explorers are independent to each other they can be executed independently, which makes a basis for efficient utilization in the Grid. The program CICADA was used for the exploration of conformational space of single amino acids, small peptides (Leu- and Met-enkephalin), DNA and RNA fragments and also mono- and oligosaccharides. We would like to show the efficiency of grid-CICADA on Leu-enkephalin and on 19 amino acids long C-end of Caseine kinase I in native and hyperphosphorylated form.
|Keywords||CICADA, PES, conformational space, folding, grid, high-throughput computing, distributed computing|
|URL for further information||http://ncbr.chemi.muni.cz; http://egee.cesnet.cz|