Research Article

iMPTCE-Hnetwork: A Multilabel Classifier for Identifying Metabolic Pathway Types of Chemicals and Enzymes with a Heterogeneous Network

Figure 4

Bar chart to illustrate the performance of the RAKEL classifiers with SVM as the basic classification algorithm under the permutation of nodes in the heterogeneous network. “All” indicates that both chemical and protein nodes are permutated; “Enzyme only” indicates that only protein nodes are permutated; “Chemical only” indicates that only chemical nodes are permutated; “No permutation” indicates no permutation is done for chemical and protein nodes (i.e., the proposed classifier, iMPTCE-Hnetwork). (a) Accuracy; (b) exact match; (c) hamming loss; (d) integrated score.
(a)
(b)
(c)
(d)