Fig. 3

Scalable application architecture. For effective use of time and space, this manuscript proposes a scalable method. Where query and reference databases are split into M and N subsets respectively. Unique pairs of query and database subsets were generated. An array job of M x N pseudo-parallel tasks takes place on the HPC cluster produces partial result for its unique pair by every task. After completing all the tasks, all the generated partial results were aggregated or merged