An Upper Bound on the Power of DNA to Distinguish Pedigree Relationships

<b>Background/Objectives:</b> Dense genetic marker panels are increasingly used in kinship analysis for the identification of distant relatives. As more markers are available, it is possible to pinpoint IBD segments more precisely and more reliably, ultimately approaching close to continuously observed IBD. This study investigates the evidential value obtained for discrimination between common pedigree relationships if IBD is observed continuously across the autosomal genome without error. In the continuous case, the evidential value is limited only by the pedigree relationship and the recombination rates. <b>Methods:</b> We conducted simulations to generate IBD segments across the autosomal genome for individuals with defined pedigree relationships. The evidential value for relationship discrimination was then calculated exactly from the underlying model, assuming no genotyping error and full genome coverage. <b>Results:</b> The simulations show that the ability to distinguish pedigree relationships quickly diminishes as relationships become more distant. First cousins can be distinguished from second cousins with 99.9% accuracy which drops to 94% when distinguishing second and third cousins. Relationships with the same expected degree of relatedness can be discriminated using continuously observed IBD, although the effectiveness decreases with more distant relationships. <b>Conclusions:</b> Continuous IBD observation establishes a theoretical upper bound on the power to distinguish relationships if a large but finite number of markers is used. The findings provide a benchmark for evaluating kinship analyses based on finite genetic marker panels.