Conformational B-Cell Epitopes Prediction from Sequences Using Cost-Sensitive Ensemble Classifiers and Spatial Clustering

Table 8

Performance of different clustering algorithms on benchmark datasets.

Method	Bound dataset		Unbound dataset
	𝑉site	𝑉𝑝	𝑉site	𝑉𝑝
PAM^1	41%	46%	40%	41%
CLARANS^2	45%	48%	48%	51%
Our method^3	53%	55%	52%	53%

1dist = Euclidean distance, 𝑘 = 4; ²dist = Euclidean distance, 𝑘 = 3; ³𝑅_avg = 19, 𝛼 = 1.1.