Skip to main content

Table 4 Sensitivity and specificity of classifications on applying rule to benchmark dataset

From: A novel strategy for classifying the output from an in silicovaccine discovery pipeline for eukaryotic pathogens using machine learning algorithms

Rule description

SN

SP

NO if TMHMM_AA < 12 and WoLF PSORT < 15 else YES

0.43

0.97

NO if TMHMM_TM = 0 and WoLF PSORT < 15 else YES

0.41

0.97

NO if Phobius_TM = 0 and WoLF PSORT < 15 else YES

0.41

0.90

NO if TMHMM_TM = 0 and MHCI < 0.5 else YES

0.63

0.84

NO if Phobius_TM = 0 and MHCII < 0.5 else YES

0.46

0.80

NO if TMHMM_AA < 18 and TargetP_SP < = 0.55 else YES

0.39

1.00

NO if TMHMM_TM = 0 and Target_SP < 0.55 else YES

0.31

1.00

NO if Phobius_TM = 0 and TargetP_SP < 0.45 else YES

0.34

0.93

NO if TMHMM_TM = 0 and SignalP < 3.8 else YES

0.24

1.00

NO if TMHMM_AA < 10 and SignalP < = 0.38 else YES

0.26

1.00

NO if TMHMM_AA < 12 and Phobius_SP = ‘N’ else YES

0.31

0.96

NO if TMHMM_TM = 0 and Phobius_SP = ‘N’ else YES

0.29

0.96

NO if TMHMM_AA < 18 and TargetP_SP < = 0.55 and MHCI < 0.5 else YES

0.31

0.84

NO if Phobius_TM = 0 and SignalP <0.45 else YES

0.21

0.93

NO if Phobius_TM = 0 and Phobius_SP = ‘N’ else YES

0.24

0.89

NO if TMHMM_AA < 18 and TargetP_SP < = 0.55 and WoLF_PSORT_annotation = NOT_screted_or_membrane else YES

0.37

0.73

NO if TMHMM_AA < 18 and TargetP_SP < = 0.55 and MHCII < 0.5 else YES

0.24

0.84

  1. Abbreviations: SN = sensitivity; SP = specificity.
  2. Note: In benchmark dataset, number of YES classifications = 70; number of NO classifications = 70; total number = 140.