Figure 2

Relationship between the log-likelihood score and the coverage of different data sources. The relationship between log-likelihood score and coverage (percentage of P. infestans proteome) of the different data sources is displayed. Projected physical interactions are shown by dots and for BioGRID (number of supporting pubmed entries (PM), e.g. 1PM or 3PM) and IntAct (core or attachment) further subdivided based on the reliability. Continuous data sources, e.g. co-expression, are indicated as a line. The dashed red line shows the applied cutoff to include associations between proteins in the predicted P. infestans network. The coverage and log-likelihood score for the P. infestans association network is denoted with a red star.