Figure 6
From: Coverage and error models of protein-protein interaction data by directed graph analysis
Geometric visualization of the solution curves from the algebraic equations 1 to 3. (a) Plot of (PFP,PFN) parameterized by n for the affinity purification-mass spectrometry (AP-MS) datasets. (b) Curves for the yeast two-hybrid (Y2H) datasets. (c) AP-MS data filtered for the proteins that were rejected by the binomial test for systematic bias. (d) curves for the Y2H data with the application of the analogous filters. These curves give upper bounds for the values of (PFP,PFN) in the multinomial error model for each experiment. Each point on any of the curves represents three distinct values based on the methods of moments restricted to the viable bait/prey (VBP) proteins: the true number of interactions between the VBP proteins, the PFP rate, and the PFN rate. If one of these three parameters can be estimated, then the other two will also be determined.