Fig. 7
From: GNE: a deep learning framework for gene network inference by aggregating biological information
The percentage of true interactions from GNE’s predictions with different probability bins. a yeast b E. coli. We divide the gene pairs based on their predicted probabilities to different probability ranges (as shown in the x-axis) and identify the number of predicted true interactions in each range. Each bar indicates the percentage of true interactions out of predicted gene pairs in that probability range