Discriminating between different modes of inheritance in genetic disease.

Abstract

A genetics program library (Morton 1969a) was used to study the problem of discrimination between different modes of inheritance in genetic disease. Data were generated on familial frequencies and on the distribution of affected sibships by a two‐allele single‐locus model and the goodness of fit was tested hy a multifactorial model to the data and vice versa.The singlelocus model is very flexible and can f i t multifactonal data well except when the frequency is very low and the heritability is high. The multifactorial model gives a poor fit in simple Mendelian‐like situations, but the fit improves as the parameters of the singlelocus model become less Mendelian. In general it will be difficult to discriminate between models of inheritance with the types of data and forms of analyses studied. Even segregation analysis does not seem critical in discrimination, but ra!her serves to confirm the results derived from the familial frequency data. In practice, with sampling errors, ascertainment biases, mortality, variable onset age, heterogeneity, and many other complicating factors, discrimination between different modes of inheritance is likely to be very difficult indeed.