Table 2 Predictions more correlated for rare than for common 1KG SAVs*.

From: Common sequence variants affect molecular function more than rare variants?

  SNAP2 CADD PolyPhen-2 SIFT
SNAP2   0.36 ± 6.8*10−3 0.31 ± 6.9*10−3 0.41 ± 6.6*10−3
CADD 0.57 ± 1.8*10−3   0.70 ± 5.2*10−3 0.63 ± 5.7*10−3
PolyPhen-2 0.53 ± 1.9*10−3 0.77 ± 1.6*10−3   0.47 ± 6.4*10−3
SIFT 0.57 ± 1.8*10−3 0.65 ± 1.7*10−3 0.49 ± 1.9*10−3  
  1. *Pearson correlation coefficients above the diagonal show the agreement for common 1KG SAVs, those below the diagonal for rare 1KG SAVs. Correlation values were calculated using the predicted raw scores of SAVs for which predictions were available for each method. The correlation in predictions between the four methods was higher for all pairs below the diagonal, i.e. for rare SAVs. Standard error of r: SEr = sqrt((1 − r2)/(n − 2)); ncommon = 18,876; nrare = 209,928.