Skip to main content
. 2022 Nov 21;109(12):2163–2177. doi: 10.1016/j.ajhg.2022.10.013

Table 3.

Assessment of the strength of evidence for PP3 through local likelihood ratios provided by three commonly used tools using default developer-recommended thresholds

Method Pathogenicity score threshold lr+ Evidence strength Fraction of predicted pathogenic variants in the gnomAD set
SIFT 0.050 0.048 (0.025, 0.075) not met 0.504
PolyPhen-2 0.902 1.993 (1.560, 2.492) not met 0.293
CADD 20.000 0.157 (0.107, 0.215) not met 0.651

Numbers in parentheses indicate lower and upper (95%) confidence intervals as calculated by the bootstrapping method. “Evidence strength” was determined on the basis of the point estimate of the positive likelihood ratio. The Pearson correlation between outputs of SIFT and PolyPhen-2 on our gnomAD set was 0.47, the correlation between SIFT and CADD was 0.49, and the correlation between PolyPhen-2 and CADD was 0.67. Correlation between other tools is provided in Figure S1.