Figure 4.

Overall methodology for predicting the functional annotations of a test protein using the label similarity-incorporated k-NN algorithm. This methodology is used for our cross-validation experiments, as well as making predictions for test proteins not annotated with any of the target functional classes.

Pandey et al. BMC Bioinformatics 2009 10:142   doi:10.1186/1471-2105-10-142
Download authors' original image