Fig. 5: Comparison of REVEL and AlphaMissense in silico tools.
From: Refinement of the classification of DDX41 variants through analysis of aggregated clinical datasets
The ability to classify putative pathogenic (n = 61) and non-pathogenic (n = 503) variants, based on the presence of any concurrent single recurrent somatic variant, was evaluated and compared. A Receiver Operating Characteristic (ROC) curve based on REVEL scores. B ROC curve based on AlphaMissense scores. C Sankey diagram illustrating the classification of variants as pathogenic supporting, benign supporting, or not met, based on REVEL (≥0.7 and ≤0.3) and AlphaMissense scores (≥0.792 and ≤0.169), across 255 evaluable missense variants with varying levels of odds of pathogenicity (OddsPath, based on the presence of somatic DDX41 variant). Six missense variants are excluded due to missing HGVSc information (n = 5) or delins variant type (n = 1). Note that this is not the final PP3 or BP4 classification, which also considers the potential splicing impact (e.g., by SpliceAI score).
