Table 1 Precision, Sensitivity, and F1 Scores from Random Forest Species Classification.

From: Utilisation of Ambient Laser Desorption Ionisation Mass Spectrometry (ALDI-MS) Improves Lipid-Based Microbial Species Level Identification

Species

Hand-Held Bipolar

High-Throughput Monopolar

ALDI-MS

50 to 500

600 to 1000

50 to 500

600 to 1000

50 to 500

600 to 1000

CALB

97

100

100

100

100

100

CDIF

97

100

100

100

100

100

ECOL

94

94

79

97

97

100

HINF

100

100

83

100

100

100

KPNE

100

97

88

97

97

97

LACJ

90

100

84

97

100

100

PAER

100

100

100

100

100

100

PMIR

93

97

90

100

97

100

SAUR

100

100

97

100

97

97

SPNE

90

100

73

97

100

100

Average

96.1

98.8

89.4

98.8

98.8

99.4

  1. The species-level classification accuracy using Random Forest for REIMS using electrical diathermy and ALDI-MS thermal disruption mechanisms for ten species is shown. For each species and thermal disruption mechanism, the classification accuracy score for mass ranges of 50 to 500 m/z (fatty acids and metabolites) and 600 to 1000 m/z (complex lipids) are shown. Full taxonomic names for each species abbreviation are given in the supporting information.