As a part of our ongoing effort to update our variant effect predictors with calibration for ACMG/AMP classifications, we have updated the Variant Effect Predictors section of our Single Variant Page. We have simplified the display of the effect predictors to make them easier to interpret. The raw score of the effect predictor is displayed, but rather than the hard to interpret and sometime unclear predictions of the individual effect predictors, we will display the calibrated BP4 Benignity and the PP4 Pathogenicity classifications.
Additionally, we have removed the circle graph that summarized the predictions. That graph relied on a relatively subjective grouping based on the text of the predictions of individual Effect Predictors. Further, it promoted a search for consensus among Variant Effect Predictors without accounting for correlation among the annotators. We are in the process of designing a view to show the correlation among the calibrated Effect Predictors to help users better evaluate the similarities and differences among Effect Predictors.
We will continue adding Variant Effect Predictors as we complete the calibrations for them. Read more about our efforts to incorporate Calibrated Variant Effect Predictors into OpenCRAVAT.
