mars-wei / molecular-graph-bert

semi-supervised learning for molecular property prediction

Molecular-graph-BERT

semi-supervised learning for molecular property prediction

requried package: tensorflow==2.3.0,rdkit==2020.03.2,numpy==1.18.5,pandas==1.1.0, openbabel==2.3.1

-- pretrain: contains the codes for masked atom prediction pre-training task.

-- classification and regression: contain the code for fune-tuning on specified tasks

-- dataset: contain the code to building dataset for pre-traing and fine-tuning

-- utils: contain the code to convert molecules to graphs

--data: data used for pretraining and fine-tuning

User should first unzip the data file and place it in the right place. Then pre-training the MG-BERT for 10~20 epoch. After that, the classification or the regression file is used to predict specific molecular property.