sacdallago / bio_embeddings

Get protein embeddings from protein sequences
http://docs.bioembeddings.com
MIT License
460 stars 65 forks source link

detokenize/reverse embedded array #199

Open drorhunvural opened 2 years ago

drorhunvural commented 2 years ago

After using ProtTransBertBFDEmbedder() or any other embedding methods, is there any way to detokenize/reverse embedded array?

For example,

embedder_bertbfd = ProtTransBertBFDEmbedder()

embedding = embedder_bertbfd .embed(sequence)

reduce_per_protein = embedder_bertbfd .reduce_per_protein(embedding)

Is there a way to get the original sequence again by using 'reduce_per_protein'?

dMedinaO commented 1 year ago

Hi everyone,

I am also very interesting in this point!

Best