Table 1.
A List of Current miR Target Prediction Tools.
Figure 1.
Three steps in the proposed semi-automated ontology development: (i) develop a backbone ontology; (ii) align the backbone ontology with other ontologies/schemas; and (iii) augment the backbone ontology.
Figure 2.
The development of a backbone ontology.
Table 2.
Sample Concepts in the Backbone Ontology.
Table 3.
Sample Relationships in the Backbone Ontology.
Figure 3.
Weight convergence experimental results when aligning TarBase with the backbone ontology, where was set to 0.1 in (a) and 0.3 in (b), respectively.
Table 4.
Characteristics of Test Ontologies/Schemas.
Table 5.
Pairwise Alignment Results among Four Ontologies/Schemas.
Figure 4.
A screenshot from Protégé, demonstrating the concept miRNA and its parents, ancestors, descendants, and siblings in is_a hierarchy.
Figure 5.
A screenshot from OBO-Edit, demonstrating more details of parents, ancestors, and direct descendants of the concept miRNA.
All relationships exhibited in this figure are is_a relationships.
Figure 6.
Another OBO-Edit screenshot, demonstrating a subset of relationships designed for the concept miRNA.
Many of these relationships are miR domain-dependent ones.
Figure 7.
A two-layer, ANN designed for the learning problem.
Figure 8.
Pseudocode 1 — ANN Weight Learning.
Figure 9.
Pseudocode 2 — Agglomerative Clustering.