Fig. 6
From: BioBLP: a modular framework for learning on multimodal biomedical knowledge graphs
Classification performance for Random Forest Classifiers for DPI benchmark with positive-negative ratio of 1:10. The box for the baseline classifier using Random embeddings is omitted for clarity, reaching mean AUPRC of 0.192 and mean AUROC of 0.720. The median is indicated by the horizontal line, the quartiles indicated by the box, the min and max by the whiskers and individual points are outliers