I have two proposals, not sure which ones are within the scope of this:
1) Compare protein language models as well in addition to small mol SMILES?
(or if you want to keep it to small mols)
2) use the gzip representations in the input embeddings for ligand docking ML models to see if they beat SOTA.
Tanmoy Sanyal, PhD
Protein design scientist at Novo Nordisk Research Seattle
I have two proposals, not sure which ones are within the scope of this: 1) Compare protein language models as well in addition to small mol SMILES?
(or if you want to keep it to small mols) 2) use the gzip representations in the input embeddings for ligand docking ML models to see if they beat SOTA.
Tanmoy Sanyal, PhD Protein design scientist at Novo Nordisk Research Seattle