Table 2 Cross-Validation of REIMS Modality Classification Models.

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

Model →  ↓Test

50 to 500 m/z

600 to 1000 m/z

ALDI-MS

Bipolar REIMS

Monopolar REIMS

ALDI-MS

Bipolar REIMS

Monopolar REIMS

ALDI-MS

N/A

82%

41%

N/A

97%

85%

Bipolar REIMS

74%

N/A

38%

88%

N/A

86%

Monopolar REIMS

51%

62%

N/A

97%

93%

N/A

  1. To determine the similarity of speciation classification models created using Random Forest analysis of REIMS mass spectra acquired using the three REIMS modalities described. Here, a speciation classification model was created using training data for both mass ranges for each REIMS modality and the processed mass spectral data from the remaining two modalities used as a test set for the model. The resulting classification accuracy for each of the 12 comparisons shows high similarity between all three modalities within the complex lipid region (600 to 1000 m/z), but low similarity within the fatty acid and metabolite region (50 to 500 m/z).
Back to article page