pykao/molbert repository overview

This docker image builds the MolBERT dictionary in numpy format for the given molecules.

Please save your molecules into a single smiles.csv file with only one column.

The title of this column should be SMILES and each row only contain one SMILES.

SMILES                                                                                              
C(=O)NCCCS(O)(=O)=O                                                                                 
CC(=O)NC1=CC=C(O)C=C1                                                                               
NC12CC3CC(CC(C3)C1)C2                                                                               
CC[N+](CC)(CCNC(=O)C(=O)NCC[N+](CC)(CC)CC1=CC=CC=C1Cl)CC1=CC=CC=C1Cl                                
CN(C)CCC=C1C2=CC=CC=C2CCC2=CC=CC=C12                                                                
CC(N)CC1=CC=CC=C1                                                                                   

Then run the following command to build the molbert dictionary for the given smiles.csv.

docker run --rm -v </local/directory/of/smiles.csv>:/data pykao/molbert --input /data/smiles.csv

After running the above command, you are able to find the resulting smiles_molbert_dict.npy in the same directory as smiles.csv

You are able to load this MolBERT dictionary using:

molbert_dictionary = np.load('smiles_molbert_dict.npy',allow_pickle='TRUE').item()

Please kindly cite our work if you find this repository is helpful to you.

1. Kao, Po-Yu, et al. "Toward Drug-Target Interaction Prediction via Ensemble Modeling and Transfer Learning." 2021 IEEE International Conference on Bioinformatics and Biomedicine (BIBM). IEEE, 2021.

ref: Fabian, Benedek, et al. "Molecular representation learning with language models and domain-relevant auxiliary tasks." arXiv preprint arXiv:2011.13230 (2020).