The use of Gene Ontology terms for predicting highly-connected 'hub' nodes in protein-protein interaction networks

Table 3 Prediction performance of the hub classifier in the combined data set of the four species

Hub classifier (# of nodes in each tree = 15, FN: FP penalty = 1:1.9, total # of trees = 187)
Training
observed	predicted non-hub	predicted hub	sensitivity	specificity	accuracy	PPV	NPV
non-hub	13381	1405	36.51%	90.50%	85.37%	28.75%	93.14%
hub	986	567
Testing
observed	predicted non-hub	predicted hub	sensitivity	specificity	accuracy	PPV	NPV
non-hub	4415	514	28.10%	89.57%	83.75%	22.00%	92.25%
hub	371	145
All
observed	predicted non-hub	predicted hub	sensitivity	specificity	accuracy	PPV	NPV
non-hub	17796	1919	34.41%	90.27%	84.96%	27.06%	92.91%
hub	1357	712

The observed vs. predicted hubs and non-hubs and their corresponding classification statistics are shown for the best classifier based on the training, testing and all (training + testing) data sets

ISSN: 1752-0509