This is the code for the paper DELE: Deductive $\mathcal{EL++}$ Embeddings for Knowledge Base Completion.

Repository Overview

• run.py is an example of how to train and evaluate a model
• evaluation_utils.py: rank-based evaluator
• losses/elembeddings_losses.py: GCI loss functions for ELEmbeddings model
• losses/elbe_losses.py: GCI loss functions for ELBE model
• losses/box2el_losses.py: GCI loss functions for Box2EL model
• data_utils/datasets.py dataset classes
• data_utils/dataloader_go.py: ontology dataloader for GO & STRING data
• data_utils/dataloader_owl2vec_star.py: ontology dataloader for OWL2Vec* data
• data_utils/deductive_closure.py: deductive closure computation
• models/elembeddings_go.py: ELEmbeddings model for PPI and function prediction
• models/elembeddings_owl2vec_star.py: ELEmbeddings model for FoodOn data
• models/elbe_go.py: ELBE model for PPI and function prediction
• models/elbe_owl2vec_star.py: ELBE model for FoodOn data
• models/box2el_go.py: Box2EL model for PPI and function prediction
• models/box2el_owl2vec_star.py: Box2EL model for FoodOn data
• data/food_ontology: contains train/validation/test sets for FoodOn
• data/go_string/*_yeast.owl: train/validation/test sets for PPI prediction
• data/go_string/*_yeast_hf.owl: train/validation/test sets for protein function prediction

Dependencies

• Python 3.9
• Anaconda

Set up environment

git clone https://github.com/bio-ontology-research-group/DELE.git
cd DELE
conda env create -f environment.yml
conda activate dele_env