Skip to main content

Table 11 Experimental results using six existing TIS predictors on our transcript collections

From: Representative transcript sets for evaluating a translational initiation sites predictor

Data Set

Approach

Sn

Sp

AA

OA

H. sap.

NetStart

83.59%

69.35%

76.47%

70.00%

 

TISHunter

94.99%

99.79%

97.39%

99.57%

 

CUB

80.14%

99.04%

89.54%

98.17%

 

StartScan

96.72%

61.29%

79.01%

62.92%

 

GENSCAN

79.02%

99.11%

89.07%

98.19%

 

GlimmerHMM

68.29%

98.39%

83.34%

96.94%

M. mus.

NetStart

83.43%

68.00%

75.71%

68.70%

 

TISHunter

70.42%

98.76%

84.59%

97.48%

 

CUB

79.11%

99.01%

89.06%

98.11%

 

GENSCAN

84.07%

99.35%

91.71%

98.66%

C. ele.

CUB

81.30%

99.28%

90.29%

98.62%

D. mel.

CUB

83.65%

99.19%

91.42%

98.45%

C. ele. II

CUB

79.99%

99.19%

89.59%

98.45%

  1. Four metrics are employed in our experiments: sensitivity (Sn), specificity (Sp), adjusted accuracy (AA) and overall accuracy (OA).