Figure 2.
Prediction quality classes and decision scores. The different prediction quality classes are shown in the coordinate system spanned by the score component Sppt and the total score S. It should be noted that Sppt is never positive for real sequences due to the construction of the mathematical expression involved (Eisenhaber et al., 1999). “P” is considered predicted; “S” denotes twilight zone predictions. With scores in the “I” and “N” zones, sequences are not predicted to have the C-terminal GPI lipid anchor motif. For more detail, see text.