Gold-standard sets of variants were defined using orthologous data from ClinVar, cBioPortal and population sequencing. In f, all variants with at least two cBioPortal ccRCC observations or one ccRCC observation and a ClinVar ‘pathogenic’ or ‘likely pathogenic’ annotation (n = 120 ccRCC-associated SNVs) are plotted with variants deemed ‘benign’ or ‘likely benign’ in ClinVar and seen at least once in population sequencing (n = 108 neutral SNVs).