Table 2 Example of performance metrics of Class 1 systems by utilizing the classic and optimized horizontal cutoff.

From: Exploring the ability of the MD+FoldX method to predict SARS-CoV-2 antibody escape mutations using large-scale data

 

System

TP

FP

TN

FN

Precision

Recall

F-Score

Cutoff\(_{\Delta G}\)

Cutoff\(_{\Delta \Delta G}\)

Classic cutoff

 Class 1

AZD8895

22

10

838

19

0.69

0.54

0.65

− 12.7

2.0

S2E12

27

7

817

38

0.79

0.42

0.67

− 11.2

2.0

REGN10933

19

26

798

44

0.42

0.30

0.39

− 17.5

2.0

C105

3

6

816

64

0.33

0.04

0.15

− 9.5

2.0

S2H14

12

11

781

85

0.52

0.12

0.32

− 12.8

2.0

Optimized cutoff

 Class 1

AZD8895

20

6

842

21

0.77

0.49

0.69

− 12.5

2.2

S2E12

34

7

817

31

0.83

0.52

0.74

− 11.5

1.7

REGN10933

16

17

807

47

0.48

0.25

0.41

− 16.9

2.6

C105

45

64

758

22

0.41

0.67

0.45

− 11.1

0.4

S2H14

57

46

746

40

0.55

0.59

0.56

− 14.0

0.8

  1. Cutoff\(_{\Delta G}\) and Cutoff\(_{\Delta \Delta G}\) correspond to cutoffs expressed in terms of \(\Delta G\) or \(\Delta \Delta G\), respectively. TP is true positive, FP is false positive, TN is true negative, and FN is false negative.