Skip to content

Limitations to inclusion list variant prioritisation

There are several limitations to inclusion list prioritisation which may lead to prioritising non-reportable variants. These situations are detailed below and these will not be reportable downstream after variant review by a clinical scientist. Many of these variants have already been identified and have been added to the exclusion list, however this will not be comprehensive and these may still occur.

  • Variant proximity: Only variants located within 2kb of a gene in the panel can be prioritised.
  • Genes with multiple modes of inheritance: For genes annotated with multiple modes of inheritance in the Newborn Panel, heterozygous, homozygous and compound heterozygous variants will be prioritised across the gene. The pipeline does not currently support variant-specific rules. This means that if a gene is considered both as monoallelic and biallelic then variants will be prioritised regardless of the reported mode of inheritance for this specific variant in the literature. (Note: this is not the case for the LOF algorithm, this can be applied to just a single mode of inheritance).
  • Overlapping and nearby genes: Variant matching for inclusion lists is by genomics coordinates and alleles, therefore if there are pathogenic variants in the overlapping or nearby gene that are also annotated to the Newborn panel gene then these will be prioritised. This will not occur with the internal inclusion lists as only Newborn panel genes are included here but this can happen with ClinVar and CVA variants. This can result in incidental findings of prioritising pathogenic variants in a gene not included in the Newborns gene panel. Incidental findings will not be reported to study participants. A summary of the overlap of newborn genes with all other OMIM genes can be found in the overlapping genes analysis report.
  • Genes associated with multiple conditions or multiple mechanisms: For CVA and ClinVar variants on the inclusion lists, we are not able to match on the condition or mechanism a variant is associated with. This may mean that pathogenic variants relevant to other conditions or mechanisms not relevant to Newborn screening may be prioritised.

  • Variants that are only clinically relevant in combination with other variants: There can be variants that are only clinically relevant when they appear in a specific haplotype in combination with other variants. Variant prioritisation is not able to only prioritise under very specific conditions for these which means that non-reportable variants that cause a mild phenotype or are only relevant with other variants will be prioritised. The most frequent examples of this scenario that have come through our analytical validity analysis (see Appendix for details) that we have identified are:

    • Variants in the gene BTD: ENST00000643237.3:c.451G>A can occur in cis with ENST00000643237.3:c.1270G>C as part of a complex allele causing profound biotinidase deficiency. However, ENST00000643237.3:c.1270G>C has now been excluded from pipeline prioritisation due to the high number of variants that require manual review as this variant in isolation is not reportable. If ENST00000643237.3:c.451G>A is observed, further manual review to detect ENST00000643237.3:c.1270G>C will be required. Instructions for this can be found in the Variant Review SOP.
    • Variants in the gene ERCC2: chr19:45352249:G:C and chr19:45357368:G:C can be prioritised as compound heterozygous in the pipeline; however, these are typically in cis and form a complex allele, which requires a third variant to be considered pathogenic.
    • CVA does not support MNVs therefore in CVA, a pathogenic MNV can be reported as two separate SNVs or sometimes a single SNV where the clinician has commented in the notes that this is an MNV. This means that the decomposed SNVs may be prioritised separately, however during manual review the clinical scientist will have access to CVA and will be able to see the context of the variant.

Where possible, variants that could be prioritised due to limitations of the inclusion lists have been curated and added to the exclusion list. However, some may be missed and prioritised as a result of these limitations.

For inclusion lists which are utilised for prioritising variants based on protein matching, it is not currently distinguished whether a variant has been prioritised by matching a variant or by matching the predicted protein change of a variant. More information can be found in the VRT User Guide.

Protein matching is not restricted by specific transcripts. Variants that match the protein change of a variant in an inclusion list in any annotated transcript can be prioritised. This may result in a variant being prioritised even if it is synonymous on a MANE transcript.

Variants may also be prioritised on the basis of protein matching with a variant on an inclusion list even if the exact variant is annotated as benign, likely benign, or a variant of uncertain significance in an inclusion list. More information about this behaviour can be found in the VRT User Guide and Variant Review SOP.