1. ClinGen-SVI suggested to downgrade PM2 (extremely low AF) to PM2_Supporting. It was recommended to refine the classification rules to adapt to changes in PM2 (for example, PVS1+PM2_Supporting should be classified as LP), but lack of clear quantitative standard. 2. ClinGen-SVI released an exception list on variants with but having potential significance. 3. Some gene-specific guidelines provided recommended precise thresholds for PM2/BA1/BS1/BS2 [41].
PM2: absent or at an extremely low frequency in population databases.
Laboratories may use different population databases [32]. The AF thresholds may vary among laboratories [11, 39, 40, 50, 51].
BA1: AF is >5% in the population databases.
BS1: AF is greater than disease prevalence.
BS2: in fully penetrant diseases, the variants were found in healthy adults as homozygosity for AR, heterozygosity for AD, or hemizygosity for XL inheritance.
The judgement on disease penetrance may be inconsistent [11].
Computer/predictive evidence
PP3: multiple in silico algorithms predict the variant to be deleterious.
Labs may have different strategies for multiple tools [15]. Using different computer prediction software can lead to discordance [11, 15].
1. ClinGen-SVI launched calibration of computational tools for missense variant pathogenicity classification [52]. 2. Several gene-specific guidelines provided recommended in silico tools.
BP4: multiple in silico algorithms predict the variant to be benign.
BP7: a synonymous variant would not impact splicing and is predicted by splicing prediction algorithms.
Functional experimental evidence
PS3: well-established in vitro or in vivo functional studies show damaging effects on genes and gene products.
The definition of “well-established” functional experiments may vary among laboratories [11].
ClinGen-SVI proposed a decision-making workflow for evaluating PS3/BS3 [53].
BS3: well-established in vitro or in vivo functional studies show no damaging effects.
Literature/database reports
PP5: reputable reports consider the variant pathogenic.
Definition of “reputable source” may not be uniform and the references or databases may vary.
ClinGen-SVI suggested to discontinue the use of these two criteria.
BP6: reputable reports consider the variant to be benign.
Null variant
PVS1: a null variant in a disease-causing gene where the loss of function (LOF) is a deleterious mechanism.
Laboratories may have divergence in judging whether LOF is the genetic pathogenic mechanism of the gene [54]. Sometimes, variants nearing the C-terminal may impact the protein function [55].
ClinGen-SVI proposed a detailed decision-making process for PVS1 refinement based on variant type, pathogenic mechanism, variant location, and inherited pattern [54].
Missense variants/in-frame insertions or deletions
PS1: a novel variant has the same amino acid alteration as a known pathogenic variant.
The use of different databases for searches for “identified pathogenic variants” can lead to differences in results.
1. Defining weather PP2/BP1/PM1 are applicable in some gene-specific guidelines [49]. 2. The ClinGen evidence repository released expert-curated variants for applying PS1/PM5 efficiently.
PM5: a novel missense variant leading to a novel amino acid change at the same locus as a known pathogenic variant.
PP2: a missense variant in a gene where these types are often pathogenic.
The specific ranges and thresholds for the frequency words “common,” “low rate,” and “primarily” are not clear enough. For example, “major” can range from “most (>50%)” to “all” or “almost all (>90%)” [11].
BP1: a missense variant in a gene where pathogenic variants are primarily truncating.
PM1: a variant in a mutational hot spot region and well-established functional domain.
Discordance in the “mutational hot spot” and “functionally defined domains without benign variants” [11].
Cosegregation evidence
PP1: cosegregation with the disease in multiple affected family members (stronger evidence if more evidence).
It is not clear how to adjust the strength of evidence for cosegregation.
1. CSER proposed the rule on cosegregation evidence weight adjustment based on inheritance pattern and a number of affected or unaffected family members [56]. 2. Well-defined in some gene-specific criteria [41].
BS4: lack of segregation in affected family members.
Coexisting evidence
PM3: a variant detected in trans with a pathogenic variant in a recessive disorder.
The criteria are not suitable for “low penetrance variants.” Judgement of low penetrant variant is highly variable.
1. ClinGen-SVI proposed a point-based system to determine the strength of in trans observations (ACMG/AMP criterion PM3). 2. Well-defined in some gene-specific guidelines [49].
BP2: a variant found in cis with a pathogenic variant in the AD gene with full penetrance or a known pathogenic variant on the same chromosome in any pattern of inheritance.
Weight adjustment
The level of evidence may be appropriately adjusted based on the evidence collected and professional judgement.
It is not clearly specified when, how, and to what extent the adjustment will be made [15].
ClinGen-SVI proposed refined nomenclature of criterion weight adjustment, such as PP1_Moderate/strong. For PS2/PM6 (de novo evidence) and PM3 (in trans criterion), ClinGen-SVI proposed scoring systems for weight adjustment.
AF: allele frequency; SVI: the Sequence Variant Interpretation Working Group; CSER: Clinical Sequencing Exploratory Research consortium.