Skip to main content

Table 10 CASP7 results, 24 or more residue separation

From: Ab initio and template-based prediction of multi-class distance maps by two-dimensional recursive neural networks

template ID

[0,10)%

[10,20)%

[20,30)%

[30,40)%

[40,50)%

[50,60)%

[60,70)%

[70,80)%

All

proteins

29

29

16

4

7

4

3

1

93

positive pairs

13532

8254

9720

2882

2972

2898

1046

644

41948

negative pairs

1163932

685474

887304

300968

244838

353646

47386

65662

3749210

SAM_T06

9.6%

9.7%

13.9%

10.5%

15.3%

8.5%

12.6%

3.6%

11.0%

Betapro

6.8%

7.9%

11.4%

10.8%

9.5%

6.0%

7.9%

2.9%

8.6%

ProfCon

11.1%

13.3%

17.2%

15.9%

13.9%

9.7%

10.4%

8.0%

13.2%

Possum

8.2%

9.0%

11.7%

8.7%

9.8%

3.4%

6.5%

0.5%

8.9%

SVMcon

10.4%

13.2%

12.9%

8.9%

14.4%

11.1%

11.3%

5.1%

11.8%

8 AI

9.1%

10.4%

11.6%

13.3%

18.4%

15.6%

9.0%

10.4%

11.2%

8 TE

8.6%

29.0%

53.1%

72.0%

80.3%

78.9%

92.5%

75.8%

37.6%

  1. Results for the CASP7 targets: sequence separations 24 residues or more, inclusive. Comparison between our two predictors (8 AI and 8 TE ) and the predictors ranked highest at CASP7. We report F1 for the contact class, as a function of sequence similarity to the best PSI-BLAST template. Predictions are from the CASP7 web site. ProfCon is in italics because it predicted 73 out of 93 maps, hence its results are not exactly comparable. In bold is the highest F1 of all predictors, excluding ProfCon, and excluding 8 TE , which uses templates and has the highest F1 in all ranges except [0,10)% where it performs worse than some of the predictors.