Skip to main content

Table 3 Comparison of ML-Select and MUFOLD-CL

From: Decoy selection for protein structure prediction via extreme gradient boosting and ranking

#

M

ML-Select

MUFOLD-CL

  

B1

B 1−2

B 1−3

B1

B 1−2

B 1−3

1dtdb

n

11.2%

11.3%

11.7%

56.2%

71.6%

76.6%

 

p

100%

100%

100%

2.42%

2.34%

2.36%

 

s

0.26%

0.26%

0.3%

54.3%

71.2%

75.7%

1wapa

n

0.31%

0.6%

0.8%

39.4%

55.1%

60.6%

 

p

100%

100%

100%

0.51%

0.49%

0.49%

 

s

0.002%

0.003%

0.004%

39%

56.4%

61.8%

1hz6a

n

4.6%

4.6%

4.6%

43.1%

48.5%

60.6%

 

p

99.8%

99.4%

98.5%

9.3%

9.2%

9.5%

 

s

0.44%

0.44%

0.45%

44.2%

50.5%

60.6%

1tig

n

3.7%

6.2%

7.1%

46.1%

61%

66.8%

 

p

100%

100%

100%

2.1%

2.2%

2.2%

 

s

0.08%

0.14%

0.16%

49.4%

61.3%

67.3%

1dtja

n

7.5%

7.9%

8.6%

57.3%

82.1%

83.5%

 

p

100%

100%

99.6%

3.0%

3.3%

3.3%

 

s

0.23%

0.25%

0.27%

59.8%

78.7%

79.3%

1bq9

n

0.62%

1.4%

2.4%

38.1%

47.7%

61.4%

 

p

100%

95.1%

83%

0.19%

0.2%

0.23%

 

s

0.002%

0.004%

0.01%

57.8%

68.9%

76.6%

1ail

n

1.4%

3.8%

3.8%

44.2%

62.3%

63.8%

 

p

100%

92.5%

86%

0.48%

0.53%

0.53%

 

s

0.01%

0.023%

0.025%

50.9%

65.3%

66.9%

1c8ca

n

0.8%

1.0%

1.1%

46.7%

59%

64.6%

 

p

100%

99%

89.1%

2.37%

2.4%

2.41%

 

s

0.02%

0.03%

0.034%

54.8%

68.4%

74.7%

2ci2

n

0%

0%

0%

45.3%

64.5%

71.7%

 

p

0%

0%

0%

3.3%

3.4%

3.4%

 

s

0.01%

0.02%

0.03%

45.3%

61.9%

69%

1fwp

n

1.84%

4.5%

4.5%

61.3%

68%

72%

 

p

97.7%

75.4%

60.3%

0.17%

0.16%

0.16%

 

s

0.003%

0.008%

0.01%

53.6%

61%

66%

1sap

n

2.63%

2.63%

2.63%

45.7%

48.9%

52.6%

 

p

87.8%

71.7%

70.6%

6.0%

5.9%

6.0%

 

s

0.21%

0.25%

0.26%

53.4%

58%

62%

1hhp

n

12.2%

18.3%

24.2%

61.2%

65.3%

85.7%

 

p

84.2%

74.8%

68%

0.09%

0.08%

0.09%

 

s

0.012%

0.02%

0.03%

57.8%

67.5%

77.1%

2ezk

n

1.3%

1.3%

1.3%

58.4%

69.5%

78.4%

 

p

59.3%

45.6%

40.3%

1.57%

1.57%

1.6%

 

s

0.03%

0.045%

0.51%

59.5%

70.6%

78.3%

1aoy

n

0.11%

0.23%

0.29%

57.3%

75%

79.7%

 

p

92.4%

92.1%

86.8%

25.8%

25.9%

25.8%

 

s

0.03%

0.07%

0.09%

57%

74.2%

79.2%

2h5nd

n

6.8%

6.8%

6.8%

43.6%

52.7%

60.5%

 

p

94.1%

83.4%

71.4%

0.35%

0.36% The

0.37%

 

s

0.028%

0.029%

0.034%

50.7%

59.3%

66.4%

1isua

n

0.021%

0.043%

0.064%

57.4%

69.3%

72.3%

 

p

17.5%

16.8%

16.4%

7.8%

7.7%

7.7%

 

s

0.01%

0.02%

0.03%

56.8%

69.4%

72.6%

1cc5

n

0.16%

0.16%

0.16%

58.8%

69.9%

72.8%

 

p

50%

42.7%

36.5%

7.3%

7.31%

7.3%

 

s

0.022%

0.026%

0.03%

54.3%

64.5%

67.2%

1aly

n

4.12%

5.2%

6.2%

47.9%

54.1%

63.9%

 

p

42.6%

42%

41.7%

0.35%

0.33%

0.34%

 

s

0.035%

0.044%

0.054%

50.5%

60.2%

69.7%

  1. The top G1−x groups of decoys selected from each selection strategy, with x limited to 3, are analyzed. When analyzing B1−x, the top x basins are merged. The analysis lists the metrics (M): percentage of near-native decoys (n); the purity (p), which is the proportion of near-native decoys relative to the size of a group; and the relative size (s, is proportional to |Ω|) of each basin