Skip to main content


Fig. 5 | Journal of Cheminformatics

Fig. 5

From: A multi-label approach to target prediction taking ligand promiscuity into account

Fig. 5

Histogram of pairwise differences of the labels predicted for test compounds annotated against two or more target proteins. The differences in performances of MMM and SMM algorithms on predicting the 5,656 labels is shown. On the x-axis, negative (positive) value indicates that the rank position, for a target, predicted by MMM (SMM) is higher than that by SMM (MMM). On the x-axis: zero denotes that MMM and SMM returned the same rank position for a target.

Back to article page