Skip to main content

Table 4 Performance on CASP10 targets using neighbourhoods (δ)

From: A study and benchmark of DNcon: a method for protein residue-residue contact prediction using deep networks

  

Acc. Top L/10(SE)

Acc. Top L/5 (SE)

Method

d

Long

Medium

Long

Medium

DNcon (222)

1

0.580 (0.032)

0.674 (0.029)

0.526 (0.029)

0.623 (0.026)

IBGteam [DL] (305)

1

0.555 (0.036)

0.648 (0.030)

0.491 (0.033)

0.609 (0.027)

RandomForest (396)

1

0.534 (0.036)

0.671 (0.030)

0.504 (0.032)

0.628 (0.028)

RaptorX-Roll (358)

1

0.529 (0.031)

0.731 (0.025)

0.490 (0.030)

0.680 (0.024)

RandomForest (257)

1

0.526 (0.036)

0.671 (0.030)

0.484 (0.032)

0.680 (0.024)

SVM (81)

1

0.394 (0.032)

0.598 (0.033)

0.365 (0.028)

0.542 (0.029)

DNcon (222)

2

0.663 (0.032)

0.749 (0.027)

0.615 (0.029)

0.720 (0.024)

RandomForest (396)

2

0.609 (0.035)

0.734 (0.027)

0.577 (0.032)

0.705 (0.026)

IBGteam [DL] (305)

2

0.607 (0.037)

0.729 (0.027)

0.555 (0.034)

0.695 (0.025)

RaptorX-Roll (358)

2

0.606 (0.031)

0.801 (0.023)

0.540 (0.029)

0.764 (0.022)

RandomForest (257)

2

0.597 (0.035)

0.734 (0.027)

0.561 (0.031)

0.723 (0.026)

SVM (81)

2

0.484 (0.034)

0.681 (0.032)

0.451 (0.034)

0.644 (0.029)

  1. Accuracies for the top L/10 and L/5 medium and long range contact predictions for 96CASP10 targets. L is the length of the protein. Estimates for standard error are provided in parenthesises. δ is the size of the neighbourhood.