Fig. 4From: Negative sampling strategies impact the prediction of scale-free biomolecular network interactions with machine learningComparison of predictive performance between ML models trained on DDB and those trained on randomly sampled datasetsBack to article page