Skip to main content

Table 1 Indexing time and peak memory of indexing for aligners using one thread. Strobealign’s indexing time and memory depends on the syncmer density used. 1 When using syncmer density 1/5 (50-300nt datasets). 2When using syncmer density 1/7 (500nt datasets). 3SNAP could not index the rye genome (Additional file 1: Note. C).\(^4\)AccelAlign does not have a singe thread mode for indexing. Multithreaded results are displayed on a node with 20 cores. We observed it used 300–700% CPU during indexing. Boldfaced values indicate the most desirable time and memory usage for each dataset

From: Strobealign: flexible seed size enables ultra-fast and accurate read alignment

 

Drosophila

Maize

hg38

CHM13

Rye

 

Time (s)

Mem (Gb)

Time (s)

Mem (Gb)

Time (s)

Mem (Gb)

Time (s)

Mem (Gb)

Time (s)

Mem (Gb)

minimap2

9

1.3

125

9.8

167

13.0

191

13.2

421

24.8

Strobealign\(^1\)

12

3.4

176

15.3

259

31

268

31.5

598

50

Strobealign\(^2\)

9

2.3

144

15.0

210

22.2

222

22.6

545

44.4

SNAP

63

1.9

1210

38.6

1744

45.9

1826

48.0

NA\(^3\)

NA\(^3\)

BWA-MEM

120

0.2

2397

3.2

3684

4.5

3629

4.6

10,245

10.7

BWA-MEM2

99

4.2

1365

63.9

3146

90.8

3000

91.3

7092

212.2

Bowtie2

228

0.3

5629

4.1

6008

5.5

7002

5.8

24,756

20.0

AccelAlign\(^4\)

10

3.4

101

28

134

38.8

132

40

320

92.5