Once we have identified millions of SNPs throughout the genome, we need to filter them. We want to identify those SNPs that will best be suited for parentage verification. From the millions of SNPs in the genome, we have selected only 500. These SNPs will now be tested at 960 birds from across the world. These birds are in different family structures and we will use the genetic information from each bird to determine which of these SNPs give the most information. If, for example, all 960 birds look the same at a specific SNP, then we can’t use it to determine parentage. We are looking for SNPs that differ between different birds and therefore give more information.
The first set of SNPs consisted of 262 SNPs, then were reduced to 195, then to 40 SNPs. These SNP-sets were tested in the families from different species and from different breeding lines. We found that the 195-set gave the most genetic information to exclude all non-parents. It can also be used to identify a bird or sibs.