Skip to main content
. 2011 May 17;27(14):1915–1921. doi: 10.1093/bioinformatics/btr303

Table 6.

Performance comparison on real datasets between drFAST-DP, drFAST-CT, Bowtie, SOCS and PerM on 1 million randomly selected reads from three different sequencing experiments

Dataset Mapper Time (min) Map locations Reads mapped (%)
NA18507 drFAST-DP 114 189 276 027 36.8
drFAST-CT 54 149 362 540 35.6
Bowtie v = 2 21 64 092 233 27.8
Bowtie n = 2 63 202 948 323 35.2
SOCS 320 21 081 941 35.3
Mapreads 80 17 032 680 37.3
PerM 76 10 068 062 35.3

NA10847 drFAST-DP 200 667 928 813 47.1
drFAST-CT 100 512 599 230 45.9
Bowtie v = 2 84 280 928 112 38.0
Bowtie n = 2 91 270 996 634 36.0
SOCS 420 53 668 622 44.8
Mapreads 140 39 589 079 48.5
PerM 100 20 699 652 44.8

NA12156 drFAST-DP 136 491 158 791 33.5
drFAST-CT 91 440 317 111 32.5
Bowtie v = 2 99 329 916 108 25.0
Bowtie n = 2 99 318 621 596 23.7
SOCS 400 38 246 530 31.2
Mapreads 110 22 182 469 35.1
PerM 140 10 798 496 31.2

We set the error threshold to 2 bp for all aligners, except PerM, where we set the threshold to four as per the PerM developers suggested. We then removed the alignments with more than 2 bp mismatches for comparison purposes.