DNA Match and Biological Relatedness: A Practical Example of the Use of Bioinformatics
by
Ranajit Chakraborty
The University of Texas School of Public Health, Houston, Texas, USA

It is commonly said that, with the exception of identical twins, no two individuals' genomes are exactly identical. Nonetheless, by chance, at several regions of the genome different individuals may have identical genotypes. In fact, the degree of partial DNA matching is an indicator of the biological relationship between individuals. However, it is a challenging problem to infer one from the other. This presentation exemplifies that using principles of bioinformatics, population genetics, and molecular genetics, inference on biological relatedness can be made from partial DNA match of individuals. Such theory also allows us to determine how many genetic loci are to be typed in DNA matching exercises to accurately infer any specific type of kinship between individuals. Turning this question in the framework of a generalized occupancy problem yields the exact probability distribution of allele/genotype sharing between individuals of different kinship categories. The method is generalized to encompass situations where the relatives are sampled from a substructured population. Confidence interval estimates of the likelihood ratio statistics are also provided. In this sense, the most general solutions are obtained to resolve the traditional DNA forensic questions as well as the ones of determining relatedness between individuals in the context of gene mapping efforts. (Research supported by US Public Health Service research grants from the National Institutes of Health and National Institute of Justice).

Date received: November 21, 2000