Automated evaluation of quaternary structures from protein crystals
Fig 3
Comparison of assembly predictions from EPPIC and PISA on the benchmarking dataset.
On the top right, a pie chart shows the global agreement between EPPIC and PISA. On the bottom left, the confusion matrix of actual (PDB1 annotations) and predicted macromolecular sizes. Tiles colored as a fraction of each EPPIC (blue) and PISA (red) macromolecular size prediction (i.e. row normalized). On the bottom right, the agreement and precision of the methods for each PISA macromolecular size prediction. On the top left, the total number and recall for each macromolecular size in the dataset.