Table 3 The runtimes (mean ± SD of three independent runs) and the memory usage of the eight UMI clustering tools on the simulated, reference, and sample data, respectively.

From: Benchmarking UMI clustering tools for accurate detection of low-frequency variants from deep sequencing

Datasets

Tools

Time (min) – mean (SD)

Memory (GB)

Default parameters

Simulated data (1.8 M reads)

AmpUMI

0.12 (± 0.01)

0.24

--min_umi_to_keep 0

Calib

0.06 (± 0.005)

0.07

-e 2 -k 4 -m 7 -t 3

CD-HIT

1.45 (± 0.10)

0.32

-c 0.90

Du Novo

0.62 (± 0.08)

0.09

-d 1

Rainbow

0.11 (± 0.02)

0.07

-m 4

Starcode

0.14 (± 0.03)

0.25

0:50:3:3

UMICollapse

1.58 (± 0.12)

18.36

-k 1 -p 0.5

UMI-Tools

1.64 (± 0.11)

18.36

--edit-distance-threshold 1

Reference data (11.3 M reads)

AmpUMI

1.49 (± 0.15)

2.59

--min_umi_to_keep 0

Calib

10.20 (± 1.20)

1.21

-e 2 -k 4 -m 7 -t 3

CD-HIT

231.51 (± 5.30)

4.14

-c 0.90

Du Novo

37.64 (± 2.90)

6.21

-d 1

Rainbow

2.59 (± 0.30)

0.99

-m 4

Starcode

2.08 (± 0.25)

4.10

0:50:3:3

UMICollapse

44.93 (± 2.40)

28.13

-k 1 -p 0.5

UMI-Tools

43.93 (± 1.50)

27.86

--edit-distance-threshold 1

Simple data (16.4 M reads)

AmpUMI

2.29 (± 0.20)

3.95

--min_umi_to_keep 0

Calib

13.94 (± 1.50)

1.79

-e 2 -k 4 -m 7 -t 3

CD-HIT

124.52 (± 8.10)

6.12

-c 0.90

Du Novo

12.63 (± 1.10)

0.91

-d 1

Rainbow

3.55 (± 0.40)

1.52

-m 4

Starcode

3.21 (± 0.35)

6.21

0:50:3:3

UMICollapse

66.60 (± 3.80)

25.20

-k 1 -p 0.5

UMI-Tools

67.36 (± 2.30)

24.86

--edit-distance-threshold 1