Benchmark disease-related performance across tools

[joerivandervelde]

It would be worth investigating whether some algorithms perform better for disease types/categories, disease subtypes, specific diseases, organ manifestation categories, perhaps even age of onset or disease severity. This could guide the use of the best tool in specific clinical presentations. For instance, perhaps VIBE does a better job at prioritizing renal genes while PhenIX does better at neuronal genes, etc. Based on input symptoms, perhaps recommendations for the best tool could even be made in the output.