Fig. 6From: GNE: a deep learning framework for gene network inference by aggregating biological informationTemporal holdout validation in predicting new interactions. Performance is measured by the area under the ROC curve and the area under the precision-recall curve. Shown are the performance of each method based on the AUROC (a, b) and AUPR (c, d) for yeast and E. coli. The limit of the y-axis is adjusted to [0.5, 1.0] for the precision-recall curve to make the difference in performance more visible. GNE outperforms LINE and node2vecBack to article page