Skip to main content

Table 8 Added value of grammar classifiers

From: Probabilistic grammatical model for helix‐helix contact site classification

Trained for

Accessibility

vdW volume

 

Avg. prop.

Grammar

Avg. prop.

Grammar

c1

0.44

0.56(0.59)

0.58

0.41 (0.45)

c2

0.43

0.45 (0.50)

0.61

0.67(0.68)

c3

0.61

0.49 (0.55)

0.64

0.64(0.71)

c4

0.62

0.57 (0.70)

0.69

0.44 (0.55)

  1. Helix‐helix contact fragments classification in the WDG150NR dataset on the basis of amino acid property average and using the combined grammars. Average AUC ROC of the best performing classifiers, out of 3 trained grammars for each class‐property case, is shown in parentheses.
Back to article page

Algorithms for Molecular Biology

ISSN: 1748-7188