Open Collections

Abstract

Genomic sequencing frequently identifies candidate variants that may cause rare genetic diseases. Some of these variants require further study to establish sufficient evidence of causation because they are variants of uncertain significance (VUS) or reside in a gene of uncertain significance (GUS). In other cases, genomic sequencing may be errantly reported as lacking candidate variants because of the failure to recognize flags of pathogenicity, such as the potential to impact splicing. Reanalysis of negative genomic sequencing can identify variants which were previously dismissed and support the reclassification of VUS as new clinical information, gene-disease associations, and bioinformatic tools are applied. Previously, splice variants were sometimes missed because filter windows were often limited to canonical splice sites (i.e. ± 1 or 2 bp away from the exon-intron boundary). As more advanced computational tools became available, potential splice variants were more readily identified, but ideally each such variant should be studied with RNA transcript analysis to confirm an impact. Using two in silico predictor tools, SpliceAI and TraP, variants from a large pediatric sequencing cohort were reanalyzed. A previously unreported candidate splice variant, GRIN2B:NM_000834.5:c.1781-19T>A, was identified in 1 of 90 undiagnosed families, introducing a possible diagnosis of GRIN2B-neurodevelopmental disorder. By applying these tools to reported variants from this cohort, it was determined that 1 in at least every 21 genomic tests may benefit from follow-up RNA-sequencing, a volume estimate which can be used to inform the implementation of the test in clinical diagnostic labs in BC. In another undiagnosed rare disease cohort, we found evidence of pathogenicity for a maternally inherited structural inversion variant in RNA-binding protein with multiple splicing (RBPMS), associating with congenital heart defects and optic nerve hypoplasia. Using segregation analysis and evidence from animal models, the case for disease causation from loss of function (LoF) variants in RBPMS was developed. All affected family members carried the inversion variant. Evidence from murine models studied by collaborators showed cardiac defects and a reduction in optic nerve size. Although efforts to identify other affected families were fruitless, RBPMS remains a strong candidate gene to cause a novel congenital malformation syndrome.