Line Plot from Scientific Research

Expected cost versus real cost. The data are all complete genomes of archaea and bacteria from the RefSeq database. The relative expected cost was computed with our model, while the relative real cost was measured with a (t_max) -HIBF with (p_fpr=0.015) , 4 hash functions and (24, 20)-minimizer. The cost is given as a ratio of expected cost to real cost for a 64-HIBF because measurements are platform-dependent. The correlation is > 0.9 , although we systematically overestimate the required time. Nevertheless, the three best (t_max) are in both cases 192, 256, and 512 with predicted values 0.53, 0.56, and 0.75 and real values 0.39, 0.42, and 0.40

DOI