Skip to main content
Figure 1 | BMC Bioinformatics

Figure 1

From: The Protein-Protein Interaction tasks of BioCreative III: classification/ranking of articles and linking bio-ontology concepts to full text

Figure 1

IMT data set class distribution. Pie charts illustrating the most frequent methods encountered in the three IMT data collections. Classes are ordered by their frequency in the test set. The most frequent training and development set classes are shown in shades of brown that in total contribute more than 50% of all class assignments in those two sets. In blue is class MI:0114 (x-ray crystallography) that is not frequent in the test set. In green are classes that are significantly more frequent in the test set than the others. MI:0018 (two hybrid) and MI:0114 are frequent in the training and development set, while MI:0416 (fluorescence microscopy) and MI:0019 (coimmunoprecipitation) are frequent in the test set.

Back to article page