RNAcentral/SO_Node2Vec

Sequence Ontology Node2Vec

This is a simple application of the Node2Vec approach to the Sequence Ontology graph for only ncRNA.

This allows us to calculate the similarity between SO types:

e.g. SO:0001877 (lncRNA) should be very similar to SO:0001463 (lincRNA) but very different to SO:0000651 (cytosolicLSUrRNA)

This SO type similarity model is used in our transcripts to genes modelling to compare the SO types of transcripts we are trying to group into genes.