Table 1 Number of called mutations (number of SNVs and indels given in brackets) and artefacts, sensitivity (sens), positive predictive value (PPV) and the F1 score regarding the eight variant calling tools investigated in case of the first and second dataset.

From: Evaluating Variant Calling Tools for Non-Matched Next-Generation Sequencing Data

Variant caller

First real dataset (HiSeq)

Second real dataset (NextSeq)

Mutations (SNVs + indels)

Artefacts

Sens

PPV

F1 score

Mutations (SNVs + indels)

Artefacts

Sens

PPV

F1 score

GATK

94 (76 + 18)

55

0.83

0.63

0.72

156 (112 + 44)

209

0.69

0.43

0.53

Platypus

95 (76 + 19)

67

0.84

0.58

0.69

189 (136 + 53)

791

0.83

0.19

0.31

VarScan

85 (69 + 15)

7

0.75

0.92

0.83

91 (84 + 7)

109

0.40

0.46

0.43

LoFreq

111 (91 + 20)

503

0.98

0.18

0.31

178 (121 + 57)

1,650

0.79

0.10

0.17

FreeBayes

113 (93 + 20)

7,918

1.00

0.01

0.03

178 (162 + 16)

35,126

0.79

0.01

0.01

SNVer

94 (76 + 18)

24

0.83

0.80

0.81

127 (72 + 55)

4,389

0.56

0.03

0.05

SAMtools

72 (70 + 2)

37

0.64

0.65

0.65

73 (71 + 2)

111

0.32

0.40

0.36

VarDict

109 (90 + 19)

9

0.96

0.92

0.94

205 (155 + 50)

3,517

0.91

0.06

0.10

Ground truth

113 (93 + 20)

0

   

226 (168 + 58)

0