From: The Protein-Protein Interaction tasks of BioCreative III: classification/ranking of articles and linking bio-ontology concepts to full text
Data set
Tot. articles
PPI
not PPI
Perc. PPI
Years
Journals
Training
2,280
1,140
50%
2007-2010
118
Development
4,000
682
3,318
17.05%
2009-2010
113
Test
6,000
910
5,090
15.00%
112
Total
12,280
2,732
9,548
-
121