Protein-protein interactions (PPIs) form the basis of biological functions. While the complete human interactome has been estimated to contain 130,000 to 650,000 binary interactions, the number of known interactions is about 30,000 to 40,000. In order to fill this gap, experimental and computational methods have been proposed to predict interactions, giving thousands of hypothesized PPIs. Studying every one of these hypothesized interactions with small-scale experiments is infeasible, as they are expensive and time-consuming.
In this talk, we discuss a collaborative web resource for studying human PPIs. Each interaction is given a predicted score of biomedical impact, based on citation behavior of past publications. In addition, users are able to describe and discuss each interaction. This feedback loop between the web resource and the user can be harnessed to identify the top K interactions that should be urgently studied because of their potential impact on biomedicine.