Fig. 2.

MBMDR-4.2.2 parallel workflow. First, each cluster node performs a fair proportion of the T _0,1,…,T _0,m values from Fig. 1 and saves the n highest into file top_c.txt. Second, a node aggregates all top_c.txt files and retrieves the overall n highest values, saved in topfile.txt. Third, each cluster node reads topfile.txt and performs an equitable fraction of the B permutations of Fig. 1, saving results into file permut_c.txt. Finally, a cluster node aggregates all permut_c.txt and produces the final output file