Against this background, we note this is a two step procedure. In the first step, the SNP data is used with both parents missing, so that the analysis depends on external allele frequency estimates and is sensitive to population stratification. In the second step, individual genotyping is performed on the parents for the top SNPs from step one, so this step is a traditional TDT and insensitive to the potential biases in step one. Accordingly, having available family DNA is needed to avoid false positive results.
