Figure 2

Comparison of literature similarity with other metrics. Literature-based similarity boxplot against evaluation metrics. X-axis corresponds to groups of biological process pairs categorized by their literature-based similarity (shown in increasing binned values). Y-axis corresponds to ontology-based similarity computed as a) Lin similarity, b) Czekanowski-Dice similarity, and c) Genome annotation similarity. Boxes expand from lower to upper quartile values, and contain a red line corresponding to median value. Dashed lines extend-ing from each end of the box show the extent of the rest of the data (with a maximum length of 1.5 units of interquartile range). Outliers, data with values beyond the ends of the whiskers, are marked as red crosses.