VEPerform: A Tool for Evaluating Variant Effect Predictors
Author Information
Author(s): Zhang Cindy, Roth Frederick P.
Primary Institution: University of Pittsburgh School of Medicine
Hypothesis
There is a need for a standardized platform to assess the performance of variant effect predictors (VEPs) at the individual gene level.
Conclusion
VEPerform enables users to evaluate the performance of different missense VEPs for individual genes, helping prioritize genes for functional testing.
Supporting Evidence
- VEPerform allows users to evaluate the performance of different missense VEPs for individual genes.
- Users can filter out common variants and customize their analysis with their own data.
- VEPerform provides both basic and advanced modes for evaluating VEPs.
Takeaway
VEPerform is a tool that helps scientists check how well different programs predict the effects of genetic changes on diseases.
Methodology
VEPerform evaluates VEP performance using balanced precision vs. recall curves based on user-selected genes and reference sets.
Potential Biases
Risk of overfitting if VEPs are trained on the reference variants used in VEPerform.
Limitations
Potential for inflated performance estimates due to circularity if VEPs are trained on the same reference set used for evaluation.
Want to read the original?
Access the complete publication on the publisher's website