Advances in high-throughput functional genomics have enabled researchers to measure many thousands of individual genetic variants in a single gene in parallel using techniques such as deep mutational scanning (Fowler and Fields, 2014). The success of these approaches depends on the availability of assays that can measure a wide range of protein functions. In their recent work, Kudla and colleagues (McDonnell et al, 2024) applied deep mutational scanning to the transcription factor PAX6, which has a key role in eye development, and described a new high-throughput functional assay that could be applied to almost any transcription factor. The authors successfully measured the effects of 95% of missense variants in PAX6 and show that their assay results are highly concordant with known clinical variants. Notably, they also undertook a wide-ranging survey of computational variant effect predictors and show that their experimental data outperformed cutting-edge algorithms.