Skip to main content

Table 1 Prediction accuracies in percent for the benchmark set. The first row indicates a prediction using relative accessible surface area in the membrane (cutoff 0.2) without lipid accessibility classification. The second row shows results for predicted lipid accessibility with our concave hull algorithm. Note that mp_lipid_acc is able to identify lipid exposed residues, giving rise to an almost 20% increase in sensitivity over a standard rASA algorithm

From: Computing structure-based lipid accessibility of membrane proteins with mp_lipid_acc in RosettaMP

 

TP

TN

FP

FN

acc

sens

spec

rASA

26526

119406

7727

10602

88.8

71.4

93.9

hull

33191

116548

10585

3937

91.2

89.4

91.7

  1. TP number of residues predicted as true positives, TN true negatives, FP false positives, FN false negatives, acc accuracy = (TP + TN)/(TP + TN + FP + FN); sens = TP/(TP + FN); spec = TN/(TN + FP)