ChEBai

ChEBai is a deep learning library that allows the combination of deep learning methods with chemical ontologies (especially ChEBI). Special attention is given to the integration of the semantic qualities of the ontology into the learning process. This is done in two different ways:

Pretraining

fit --data.class_path=chebai.preprocessing.datasets.pubchem.SWJChem --model=configs/model/electra-for-pretraining.ElectraPre --model.train_metrics=configs/metrics/micro-macro-f1.yml --model.val_metrics=configs/metrics/micro-macro-f1.yml --model.test_metrics=configs/metrics/micro-macro-f1.yml --trainer=configs/training/default_trainer.yml --trainer.callbacks=configs/training/default_callbacks.yml

Structure-based ontology extension

python -m chebai fit --config=[path-to-your-electra_chebi100-config] --trainer.callbacks=configs/training/default_callbacks.yml  --model.pretrained_checkpoint=[path-to-pretrained-model] --model.load_prefix=generator.

Fine-tuning for Toxicity prediction

python -m chebai fit --config=[path-to-your-tox21-config] --trainer.callbacks=configs/training/default_callbacks.yml  --model.pretrained_checkpoint=[path-to-pretrained-model] --model.load_prefix=generator.

python -m chebai train --config=[path-to-your-tox21-config] --trainer.callbacks=configs/training/default_callbacks.yml  --ckpt_path=[path-to-model-with-ontology-pretraining]