Fig. 6

NaviSE peformance comparisons. a NaviSE CPU running time for different numbers of CPUs. Heatmap of the processing time for each NaviSE process for different numbers of CPUs, writen on top. Tasks parallelised by NaviSE are highlighted in bold typeface. For the SE prediction and gene annotation, running times of ROSE and HOMER on 19 CPUs are also provided. b SE prediction similarities among different software. For each cell type and histone mark, the Euler-Venn diagram with the number of commonly predicted SEs is represented on top and the comparisons among the SE ranking generated by the different software at the bottom. The rank of each SE predicted by NaviSE is colour-codified (bluer colours indicate higher positions in the rank and redder colours lower positions). Each NaviSE SE is mapped onto HOMER and ROSE SE ranking tracks in the position predicted by HOMER and ROSE for such SE, with the colour codification corresponding to the ranking predicted by NaviSE. SEs predicted by other software that are not predicted by NaviSE appear in white. Grey boxes mark the indexes for which a rank in a predictor has exhausted its number of predicted SEs in comparison to the maximum rank predicted by the three software {HOMER, ROSE, NaviSE}