Bayesian Assessment of Genetic Risk in Families with a Balanced Translocation

An important problem from the field of genetics involves the calculation of a personalized risk estimate on behalf of a heterozygous carrier of a balanced translocation. Though phenotypically normal, the carrier may be at increased risk of having a child who is mentally and physically abnormal due to an unbalanced translocation of chromosomal segments. An accurate estimate of the probability of this event is understandably desirable. Unfortunately, translocations are almost always family-specific so there is very little data that are perfectly relevant and one has to rely heavily on general risk figures derived from studies of families with similar translocations. This makes the problem particularly well suited to Bayesian analysis, which coherently combines family-specific data and a priori knowledge. However, much of the genetics counseling literature recommends an either/or approach: if the family is large enough, use family data; else, use pooled population data. In this article, we describe how uncertainty can be significantly reduced by incorporating all available information in the context of deriving a risk estimate for a hypothetical familial translocation.