The main difficulty with studying recombination events in chromosomes transmitted to the children is that the parents need to be phased. But phasing the parents directly requires genotyping the grandparents, a serious limiting factor. An alternative is to utilize nuclear families with 3 or more children genotyped. Here, in effect, the children are used to phase the parents. Specifically, if both parents and two children are genotyped, one can detect a recombination event, but it would be impossible to tell in which child the event occurred. With more than two children, by assuming that the chance that more than one child has gone through a recombination event in a small region is negligible, the uncertainty in phase is resolved by the majority rule. Fig. 3 displays an example of a recombination observed at a known hotspot in the MHC region. The recombination event between SNPs rs2532924 and rs3095089 (SNP5 and SNP6) could be deduced from data of the two parents and the three children (C1 to C3). It is clear that all three children share the maternal allele IBD for
