Reply: UNC13B and focal epilepsy

The concerns in the letter by Green et al.1 refer to a critical question—how to use minor allele frequency (MAF) in screening potential pathogenic variants and select controls from MAF databases in evaluating gene-disease association. MAF is a critical factor to be considered in pathogenic variants screening. Generally, MAF ≤ 0.005 for phenotypes of autosomal dominant (AD) inheritance and MAF ≤ 0.01 for phenotypes of autosomal recessive (AR) inheritance in general populations were set for data filtration. The MAF in the general population depends on phenotype prevalence2 that is related to phenotype severity. This is because incomplete penetrance is common in genetic disorders, especially in those with mild phenotypes. For instance, incomplete penetrance was observed in 12 of the 17 families with DEPCD5 mutations.3 Our previous study demonstrated that the overall penetrance was 70.3% in familial epilepsy caused by DEPCD5 mutations4 and were 91.9% in the familial cases caused by SCN1A mutations,5 which explained the asymptomatic carriers in general populations, even in ‘normal controls’. Therefore, the very low MAF should not be an exclusion criterion for evaluating the pathogenicity of variants such as p.Trp45X and the p.Gly882Trp in UNC13B, which appeared once in all-population and were absent in controls of gnomAD6 (also in Table 1 in Green et al.1).