Skip to main content

Table 2 Overall percentage success rates for predicting the rank of the correct parent formula (\(r = 1\), \(r \le 2\), or \(r \le 4\)) for the different data sets, for different alphabets, and for the different scoring functions. The average relative ranking position \(\left\langle {\text {RRP}} \right\rangle\) and the average number of mass peaks \(\left\langle N_{\text {peak}} \right\rangle\) is also reported. For the RECETOX data set, -ClBr indicates removal of molecules with more than three of these atoms, due to potential isotope peak interference

From: Determining the parent and associated fragment formulae in mass spectrometry via the parent subformula graph

% Success rate per

Data set

Element

Rank r

Scoring function

No. of

Total no.

\(\left\langle N_{\text {peak}} \right\rangle\)

Alphabet

\(\left\langle {\text {RRP}} \right\rangle\)

\(s_{\text {ne}}\)

\(s_{\text {LBJ}}\)

\(s_{\text {v}}\)

Spectra

Of spectra

CASMI

CHNO+

1

90.3

87.7

86.6

236

575

24.8

\(\le 2\)

93.6

93.2

91.9

\(\le 4\)

97.0

97.5

95.3

\(\left\langle {\text {RRP}} \right\rangle\)

0.016

0.020

0.027

CHNOP+

1

80.0

82.4

72.1

365

\(\le 2\)

89.9

91.2

84.7

\(\le 4\)

94.2

94.8

91.2

\(\left\langle {\text {RRP}} \right\rangle\)

0.050

0.042

0.073

CHNOPF+

1

71.2

76.3

58.9

438

\(\le 2\)

82.6

87.0

73.5

\(\le 4\)

90.0

92.0

83.8

\(\left\langle {\text {RRP}} \right\rangle\)

0.058

0.046

0.091

RECETOX

CHNO+

1

66.2

64.7

59.4

133

240

56.4

\(\le 2\)

80.1

80.1

76.4

\(\le 4\)

87.0

90.1

87.8

\(\left\langle {\text {RRP}} \right\rangle\)

0.19

0.19

0.20

CHNO+ -3ClBr

1

77.5

75.5

68.1

94

45.1

\(\le 2\)

87.0

87.0

81.5

\(\le 4\)

89.1

93.5

90.2

\(\left\langle {\text {RRP}} \right\rangle\)

0.09

0.09

0.11

ORCHID

CHNOS

1

30.4

47.8

13.0

23

23

141.0

\(\le 2\)

47.8

56.5

26.1

\(\le 4\)

56.5

87.0

30.4

\(\left\langle {\text {RRP}} \right\rangle\)

0.024

0.055

0.096