Skip to main content
. 2016 Aug 18;5:37. doi: 10.1186/s13742-016-0143-4

Table 2.

Comparison of the cumulative CPU times (hours) for RES-Scanner, REDItools and GIREMI in processing the human GM12878 dataset from pre-aligned reads to final editing sites

Chromosome RES-Scanner REDItools GIREMI
chr1 39.71 118.98 ND
chr2 44.51 128.62 ND
chr3 37.95 106.15 ND
chr4 29.83 81.61 ND
chr5 31.16 88.89 ND
chr6 28.99 85.49 ND
chr7 25.60 82.45 ND
chr8 25.38 64.02 ND
chr9 20.23 63.79 ND
chr10 23.01 71.45 ND
chr11 20.80 68.59 ND
chr12 23.33 75.51 ND
chr13 14.65 40.37 ND
chr14 16.10 51.55 ND
chr15 14.82 48.73 ND
chr16 15.64 48.15 ND
chr17 14.54 54.50 ND
chr18 13.03 35.29 ND
chr19 13.07 31.63 ND
chr20 12.19 31.04 ND
chr21 7.59 15.66 ND
chr22 7.67 21.62 ND
chrX 21.80 60.09 ND
chrY 1.62 1.94 ND
Total 503.23 1,476.12 10.87

Note: A total of 150 Gb of pre-aligned DNA reads and 9 Gb of pre-aligned RNA reads in BAM format were used as inputs for both RES-Scanner and REDItools, while 9 Gb of pre-aligned RNA reads and a list of SNVs derived from the RNA-seq data were used as inputs for GIREMI. The time for GIREMI included the cumulative CPU times of generating the SNV list from the RNA-seq data using SAMtools [40] (9.60 h) and running GIREMI (1.27 h). As GIREMI required all SNVs from the whole genome to construct the MI distribution, CPU times for individual chromosomes could not be determined. ND not determined