Skip to main content

Table 1 Table: MetaBenchmark performance metrics for ten metagenomics analysis tools sorted by precision. Average numbers for the six simulated data sets are given. Accuracy evaluated at the species level and reported as sensitivity, specificity, precision (positive predictive value), negative predictive value (NPV) and Matthew’s Correlation Coefficient (MCC). Index size reported for CLARK is the sum of the phylum (46.6GiB), genus (149.5GiB) and species (146.3GiB) indexes. Performance metrics at genus and phylum ranks can be found in Supplementary Tables 2 and 3

From: UMGAP: the Unipept MetaGenomics Analysis Pipeline

Tool

Precision

Sensitivity

Specificity

NPV

MCC

Run time

Index size

UMGAP tryptic prec.

99.70%

3.50%

99.96%

21.83%

8.62%

6.96 m

19.3 GB

Kraken

99.38%

81.34%

98.15%

59.21%

68.24%

210.86 m

198.7 GB

UMGAP max prec.

98.94%

45.87%

98.22%

33.35%

37.73%

16.47 m

132.9 GB

Kraken 2

98.15%

82.27%

94.64%

60.68%

67.27%

2.10 m

43.3 GB

UMGAP high prec.

98.11%

55.68%

96.21%

38.07%

43.33%

30.42 m

132.9 GB

Kaiju

98.02%

68.31%

95.21%

46.44%

53.15%

304.10 m

74.4 GB

UMGAP tryptic sens.

96.07%

18.03%

97.36%

24.88%

17.95%

6.10 m

19.3 GB

UMGAP high sens.

92.55%

66.70%

84.12%

46.04%

44.28%

30.32 m

132.9 GB

UMGAP max sens.

80.78%

77.73%

63.34%

58.94%

40.39%

31.12 m

132.9 GB

CLARK

71.41%

100.00%

27.87%

100.0%

44.61%

20.54 m

342.3 GB