BenignPathogenic
|
|
|
|
0.2
0.4
0.6
0.8
Heatmap of EVE scores. Positions with no EVE score are shown in white.

Reference aa: Amino acid in reference sequence.
Position: Position in amino acid sequence.
Variant aa: Amino acid variant.
EVE Score: EVE pathogenicity score.
EVE Class (75% Set): EVE classification when only classifying the 75% most confident of all possible amino acid substitutions.
Class Uncertainty: EVE classification uncertainty.
Frequency (gnomAD v2): Frequency in gnomAD v2.
ClinVar Label: Label found in ClinVar.

Statistics Summary
No PDF file specified.
Number of Labels
EVE AUC (All Labels)
EVE Accuracy (All Labels)
EVE Accuracy (75% Set)
Loading PDF…
Scatter and violin plots of EVE scores for variants seen in humans to date (variants found in ClinVar, gnomAD or the UK Biobank). The dashed lines show EVE classification boundaries for the 75% most confident class assignments (over all possible amino acid variants), which on average corresponds to ~90% classification accuracy. Only ClinVar labels with one star quality rating or higher were considered.
Performance Statistics.

Only ClinVar labels with one star quality rating or higher were considered.

The 75% set corresponds to the 75% most confident class assignments (over all possible amino acid variants).



WARNING: If the number of ClinVar labels is small, these statistics are of limited value. Aggregate performance statistics over all labels of all genes can be found here.