Fig. 4: Distribution of Maximum Common Edge Subgraph distances. | Nature Communications

Fig. 4: Distribution of Maximum Common Edge Subgraph distances.

From: Coverage bias in small molecule machine learning

Fig. 4

Myopic MCES distances were computed with threshold T = 10. For the histogram plots, bin intervals are left-open and right-closed. For example, the tenth bin in the right plot covers distances x with 9 < x ≤ 10. a The distribution of all 199,870,021 computed distances from the subsampled biomolecular structures are shown as a histogram plot and kernel density estimate. The plot is truncated after distance 160. b Histogram of the minimum myopic MCES distance for each subsampled biomolecular structure against all other biomolecular structures, truncated after distance 50. The hump is due to the employed threshold T = 10, see text for details. Source data are provided as a Source Data file.

Back to article page