Abstract
Complete uniparental isodisomy (iUPD)-the presence of two identical chromosomes in an individual that originate from only a single parental homolog-is an underestimated cause of recessive Mendelian disease in humans. Correctly identifying iUPD in an index patient is of enormous consequence to correctly counseling the family/couple, as the recurrence risk for siblings is reduced from 25% to usually <1%. In medium/large-scale NGS analyses, we found that complete iUPD can be rapidly and straightforwardly inferred from a singleton dataset (index patient only) through a simple chromosome- and genotype-filtering step in <1 min. We discuss the opportunities of iUPD detection in medium/large-scale NGS analyses by example of a case of CHRNG-associated multiple pterygium syndrome due to complete maternal iUPD. Using computer simulations for several detection thresholds, we validate and estimate sensitivity, specificity, positive (PPV), and negative predictive values (NPV) of the proposed screening method for reliable detection of complete iUPD. When screening for complete iUPD, our models suggest that a >85% proportion of homozygous calls on a single chromosome with ≥30 sufficiently interspaced called variants results in a sensitivity of 97.9% and specificity of 99.7%. The PPV is 95.1%, the NPV 99.9%. When this threshold is exceeded for a chromosome on which a patient harbors an apparently homozygous disease-associated variant, it should be sufficient cause to discuss iUPD as a plausible or probable mechanism of disease in the genetic analysis report, even when parental segregation has not (yet) been performed.