Fig. 1
From: HAPPI-2: a Comprehensive and High-quality Map of Human Annotated and Predicted Protein Interactions
![Fig. 1](http://media.springernature.com/full/springer-static/image/art%3A10.1186%2Fs12864-017-3512-1/MediaObjects/12864_2017_3512_Fig1_HTML.gif)
An illustration of how we define the “database validation through PPI rediscovery” approach. The rectangle area consisting of both the S set (in green shades) and the T set (in red shades) represents all genes taken from a curated gene set x from the MSigDB database. The lines connecting genes are PPIs all taken from a single database under validation. The G set shows new (“discovered”, outside of the rectangle) or old genes (“re-discovered”, in either the S set or the T set) by expanding from genes in the S set by one layer of PPI relationships