Genomic scanning and the transmission/disequilibrium test: analysis of error rates

A multi-allelic extension of the transmission/disequilibrium test (TDT) was applied to quantitative data from GAW10 Problem 2A by using thresholds to define affection status. Every one of 367 marker loci was screened. It was found that TDT results at neighboring loci are uncorrelated in these data (where the markers are in linkage equilibrium). It was also found that the chi-square statistics for both tests are larger than they should be so that the probability of false linkage detections is increased. Alternative approaches to the chi-square should be considered for assessing statistical significance. Results at marker D5G15 were strong, probably because of linkage without association.