The importance of terms in the physics-based scoring function is shown. A) The distributions of relative weights for each term in the full scoring function on proteins/complexes in the test sets (binding terms prefixed with "b") are depicted. Outliers are shown as empty circles. B) The gap between the native score and the distribution of random samples when using only single terms in the scoring function is depicted. Gaps are shown for the set of random sequences (dark grey), the set of random geometries (light grey) and the combined set (black). Note that only folding score terms are shown.
Grahnen et al. BMC Evolutionary Biology 2011 11:361 doi:10.1186/1471-2148-11-361