Fig. 6

Comparison of enrichment scores estimated by mutscan, DiMSum, and Enrich2. For mutscan, the values are logFCs estimated by either edgeR or limma. For DiMSum, they correspond to the enrichment score derived from all replicates. For Enrich2, they are the averages of the scores for the replicates. The lower correlations seen in the Bolognesi_TDP43_290_331 data set, and to some extent, the Diss_FOS_JUN data set reflect the less stringent variant filtering criteria used in these data sets (up to 3 mutated bases for Bolognesi_TDP43_290_331 and up to two mutated codons for Diss_FOS_JUN, see the “Methods” section). Considering only variants with fewer recorded mutations leads to higher correlations, more comparable to the Diss_FOS data set where only up to two mutated bases were allowed (Additional file 2: Figs. S11-S14)