A COMBINED TEST OF LINKAGE HETEROGENEITY

This paper focuses on the problem of testing for heterogeneity once li nkage is established. In an investigation of genetic linkage, Morton f irst proposed a general purpose test to detect heterogeneity in the re combination fraction. Two more commonly used tests of linkage heteroge neity are the admixture test (A-test) of Smith, Ott, and Risch and Bar on, and the B-test of Risch. All are likelihood-ratio tests, but they differ in the models specifying the heterogeneity. A new test of heter ogeneity in the presence of linkage is presented here. I propose a mix ture model of heterogeneity, which allows the recombination fraction t o vary among families, as does the B-model, yet also allows some famil ies to be unlinked, as the A-model does. This model contains the A and B models as special cases and thus allows a direct test (D-test), whi ch can provide justification for choosing one of these extremes.