Author: Edgard Verdura, PhD
Welcome to our February edition! This month we continue our Molecular Mechanisms series by showing how AION prioritizes variants with automated ACMG guidelines even in edge cases where frequency data could have excluded the candidate variants.
Frequency in control populations is one of the main criteria considered to determine variant pathogenicity or benignancy. However, it comes with several challenges, as thresholds to consider a variant as candidate vary depending on the variant’s molecular properties, disease prevalence and other variables, and that can complicate application of ACMG guidelines. In this newsletter we illustrate how AION assesses frequency using a whole exome sequencing case. The patient is affected by an hematological disorder; her clinical presentation includes HPOs such as “Failure to thrive” and “Anemia”. Following AION prioritization, the top ranked variant was an homozygous missense change in HBB, associated with Beta-thalassemia and Sickle Cell Anemia, among other diseases related to hemoglobin malfunction (OMIM).
This missense variant (NM_000518.5:c.20A>T; p.Glu7Val) has been tagged in ClinVar as “Conflicting interpretations of Pathogenicity” (ClinVar link). This happens because it has been described as Pathogenic by 47 different submissions (Sickle cell disease, Beta-thalassemia, others), but on the other hand, 3 submissions evaluate this variant as Likely Benign, mainly based on frequency assessment. Indeed, frequency, allele count, and homozygous individuals counts in gnomAD database are relatively high both for autosomal dominant and autosomal recessive modes of inheritance (Frequency=0.004374, gnomAD_AC=1326, gnomAD_hom=4).
Interestingly, in the African/African American subpopulation in gnomAD, the population frequency of this variant is higher than 4%! (gnomAD link). The p.Glu7Val change in HBB is well known in the field of human genetics, not only for its high prevalence, but also for providing an adaptive advantage to heterozygous carrier individuals, which are less susceptible to infection by Plasmodium falciparum (malaria). This phenomenon partly explains why this mutation is so prevalent, especially in areas where malaria is endemic (Ferreira et al., Cell 2011).
In other frameworks this variant would have been evaluated as Likely Benign or Benign by ACMG guidelines, as ACMG criteria BS1 (Allele frequency is greater than expected for disorder) or BS2 (Observed in healthy adult individuals) would be applied (Richards et al., 2015). In contrast, AION does not trigger BS1 or BS2 rules if there is strong evidence of pathogenicity from other sources, therefore prioritizing the variant effectively. While ClinGen Sequence Variant Interpretation group has recently published recommendations on how to apply BA1 (related to assessment of very high frequency variants) and its exceptions, BS1 and BS2 criteria are disease-specific, and its use depends on recommendations by expert panels for each disease, which can complicate its implementation (LINK).
Summarizing: this case shows how AION can tackle several challenges while interpreting variants, and specifically, it enables accurate assessment of pathogenicity in variants which are strongly prevalent in healthy populations.