Calculating variant penetrance using family history of disease and population data

Abstract

PurposeGenetic penetrance is the probability of a phenotype manifesting if one harbours a specific pathogenic variant. For most Mendelian genes, penetrance is incomplete and may be age-dependent. Accurate penetrance estimates are important in many biomedical fields including genetic counselling, disease research, and for gene therapy. The main methods for its estimation are limited in situations where large family pedigrees are not available, the disease is rare, late onset, or complex. MethodsHere we present a novel method for penetrance estimation in autosomal dominant phenotypes. It uses population-scale data regarding the distribution of a variant among unrelated people affected (cases) and unaffected (controls) by an associated phenotype and can be operated using samples of affected people only by considering family disease history. ResultsThe method is validated within simulation studies. Candidate variant-disease case studies yield estimates which align with existing disease knowledge and estimates derived using established methods. ConclusionWe have presented a valid method for penetrance estimation in autosomal dominant traits which avoids kinship-specific penetrance estimates and the ascertainment biases that can arise when sampling rare variants among control populations. It can be accessed from our public web server (https://adpenetrance.rosalind.kcl.ac.uk) and is available as an open-source R library (https://github.com/ThomasPSpargo/adpenetrance).