Figure 6From: Incorporating functional inter-relationships into protein function prediction algorithmsComparison of the performance of functional similarity-incorporated k -NN classifiers with individual k -NN classifiers for small classes (size ≤ 30) in Mnaimneh et al 's data set. The X-axis shows the AUC scores of the base classifiers for each class, and the Y-axis denotes the AUC scores of their label similarity-incorporated classifiers.Back to article page