We assessed genotyping error rates for all samples. Allelic
dropout was interpreted in cases where at least one amplification
trial showed a homozygous pattern while the others were heterozygous.
Alleles that occurred in only one of the PCR replicates
were interpreted as potentially false alleles. In such cases, three
additional replicates were run for the same sample to ensure that
the presumed spurious allele was indeed an amplification artefact