Skip to main content
Fig. 2 | BMC Genomics

Fig. 2

From: Predicting metabolic pathway membership with deep neural networks by integrating sequential and ontology information

Fig. 2

Generation of vector representation from GO dataset. In this example, the protein is annotated with t3 and t4. To generate protein’s feature vector, the normalized IC of t3 and t4 is used in the first stage. On second stage, the semantic value (SV) of all term ancestors of t3 and t4 are calculated. Since t3 and t4 share common ancestor, t1 and t2, the semantic value for t1 and t2 are average semantic value (\(\overline {SV} \)). See Material and Methods for detail description

Back to article page