Algorithmic re-structuring and data replication for protein structure comparison on a GRID

This paper describes a major restructuring of PROuST, a method for protein structure comparison, for an efficient porting to the Grid. PROuST consists of different components: an index-based search that produces a list of proteins that are good candidates for similarity, and a dynamic programming algorithm that aligns the target protein with each candidate protein. Both components use the same geometric properties of secondary structure elements of proteins. Thus, an important issue arises when porting the application to the Grid, i.e. the tradeoff between data transfer and data recomputation. Our restructured application avoids recomputation by re-using the data as much as possible, once they are accessed. The algorithmic changes to PROuST allow to reduce the number of data accesses to storage elements and consequently the execution time. This paper also discusses data replication policies on a Grid environment to optimize the data transfer time.