Skip to main content

Table 5 The least prediction accuracy (%) for two-class problems according to different contact number definitions. Only local sequence information is used.

From: Better prediction of protein contact number using a support vector regression analysis of amino acid sequence

  r d = 8 Å r d = 10 Å r d = 12 Å r d = 14 Å
Discrete 74.1 74.5 75.8 75.2
Consecutive 75.6 75.7 76.1 75.6