Skip to main content

Table 2 Single-feature polymorphism (SFP) discovery methods comparison

From: Single-feature polymorphism discovery by computing probe affinity shape powers

 

Methods

 

PASP

PILM

PILM

PAOP

PAOP

 

probe affinity, shape power

probe intensity, linear model

probe intensity, linear model

probe affinity, outlier score

probe affinity, outlier score

cutoff

weight2.5

D3.0 FDR0%2

D2.0 FDR0.1%1

os pct0.152

os pct0.051

polymorphisms

284/401

254/401

295/401

233/401

106/401

non-polymorphisms

48/2200

87/2200

155/2200

37/2200

16/2200

detected SFPs

9603

7150

10504

6820

2193

TPR (%)

70.82

63.34

73.56

58.10

26.43

FPR (%)

2.18

3.95

7.05

1.68

0.73

FDR (%)

14.46

25.51

34.44

13.70

13.11

compScore

1.00

0.21

0.15

0.80

0.28

  1. The PASP (probe affinity shape power) method developed in this study was compared with the PILM (probe intensity linear model) and the PAOP (probe affinity outlier pursuit) methods using the same dataset used in the previous study [23, 47]. Both the cutoff values used in original report 1 and the cutoffs at which the best performance was produced 2 in each method were tested. The 401 polymorphic and 2200 non-polymorphic sequences of Golden Promise and Morex were derived from the previously-published Barley1 GeneChip data sets [47]. The calculations for true positive rate (TPR), false positive rate (FPR), false discovery rate (FDR), detected SFPs, and compScore are described in the Materials and Methods section. CompScore is the comparison of the PASP method versus each of the other methods.