Affected sib pair identity by state analyses

Four methods using identity by state (IBS) data from affected sib pairs are compared for their ability to detect linkage between a diallelic marker and disease. A joint null hypothesis of no linkage and no linkage disequilibrium between the marker and disease must be considered. Two tests have undesirable properties in the case of linkage disequilibrium. Which of the other two tests has more power is dependent on the presence or not of linkage disequilibrium. The procedure of choice when possible is to type parents of affected sib pairs: the null hypothesis of no linkage can then be tested using identity by descent (IBD) values from informative parents, and the null hypothesis of no marker association with disease (linkage equilibrium) can be tested independently using the marker allele frequencies in the affected sib pairs.