Table 2 Criteria applied in the classification of variants

5

Clearly pathogenic

1. Reported in the literature as pathogenic with supporting evidence; multiple independent cases, pedigree segregation studies and/or functional analysis AND

2. Phenotype and inheritance pattern in patient correlates with the gene

4

Likely to be pathogenic

1. Not described in the literature, or weak evidence for pathogenicity in published literature; no segregation or functional analysis available AND

2. Phenotype and inheritance pattern in patient correlates with the gene AND

3. Location of variant in gene and pathogenic mechanism are compatible with previously described pathogenic variants in the gene AND

4. Minor allele frequency (MAF) is <1 % in dbSNP (v142) and <1 % in Exome Variant Server (EVS, v6500) AND

5. a) Missense variant; conserved amino acid, Polyphen 2 (HumVar) and SIFT concur in predicting deleterious effect or three or more of five splice prediction tools^a return >10 % difference in splice site prediction value between wild type and variant OR

b) Frameshift or nonsense variant; where expected mechanism is loss of function OR

c) Synonymous or intronic change; nucleotide highly conserved across multiple species and three or more splice prediction tools return >10 % difference in splice site prediction value between wild type and variant

3

Unknown significance (VoUS)

1. Minor allele frequency <1 % in dbSNP (v142) and EVS (v6500) AND

2. Phenotype and inheritance pattern in patient correlates with the gene AND

3. Not described in the literature, or described in literature with inconclusive or no evidence AND

a) In silico predictions score variants as Class 4 but phenotype does not correlate with the gene OR

b) In silico predictions are conflicting (for example; conserved amino acid, very low MAF but Polyphen 2 and SIFT predict benign)

2

Unlikely to be pathogenic

1. a) Minor allele frequency is between 1 % and 3 % in dbSNP (v142) and/or between 1 % and 3 % in EVS (v6500) but phenotype and/or inheritance pattern in patient correlates with the gene OR

 b) Minor allele frequency is <1 % in dbSNP and EVS but phenotype and/or inheritance pattern in patient does not correlate with gene AND

2. a) Missense variant; amino acid is weakly conserved across multiple species and/or Polyphen 2 (HumVar) and SIFT concur in predicting benign OR

 b) Synonymous or intronic change; nucleotide weakly conserved across multiple species and splice prediction tools^a show no significant difference between wild type and variant, even if MAF is <1 % on dbSNP and EVS AND

3. a) Some evidence for benign status in literature but weak or inconclusive AND/OR b) Some evidence that variant does not segregate with disease in pedigree of patient(s) tested in this cohort

1

Clearly not pathogenic

Frequency >3 % in dbSNP (v142) or >3 % in Exome Variant Server (v6500) OR

Frequency <3 % but described and proven as not pathogenic in published literature