A New Tool For Genetic Genealogists: 23andMe’s Relative Finder


23andMe has been beta testing a new tool for comparing autosomal DNA results called “Relative Finder.” Although I was not one of the beta testers, it seems that this new tool will be of great use to genealogists. Roberta Estes has posted a nice summary of the Relative Finder tool at the “Searching for the Lost Colony DNA Blog.”

90% of Customers Likely to Find a Match!

Relative Finder compares your autosomal SNP results to the results of others in the 23andMe database to determine matches. While developing the tool, 23andMe discovered that in their dataset of “more than 5000 individuals with European ancestry,” 90% of individuals had at least one distant relative between 2nd and 8th degree cousins!

Other Uses

Beyond identifying distant relatives, the tool will potentially have many other uses. For example, one individual has already used SNP testing and comparison of 6 cousins to determine the approximate chromosomal location of an autosomal dominant hereditary condition!

The ASHG Abstract

Below is 23andMe’s abstract for the upcoming 2009 ASHG (American Society of Human Genetics) meeting regarding the Relative Finder tool:

Discovering Distant Relatives within a Diverse Set of Populations Using DNA Segments Identical by Descent.

L. Hon, B.M. Henn, J.M. Macpherson, N. Eriksson, A. Wojcicki, L. Avey, S. Saxonov, J.L. Mountain. 23andMe, Inc, Mountain View, CA.

“Close familial relationships, such a parent-offspring relationships, are inferred readily from genotype and allele frequency data for a limited number of single nucleotide polymorphisms (SNPs) or short tandem repeat (STR) polymorphisms. However, researchers have not focused on accurately inferring the degree of relationship between individuals who share a common ancestor 3-10 generations ago; such inference requires high density genome-wide information not available until recently. In order to characterize relatively distant relationships for a sample, we analyzed the sharing of DNA identical by descent (IBD) in a large database of self-selected individuals of European ancestry, including a subset identified as Ashkenazim. Individuals from more endogamous ethnic populations, such as the Ashkenazim, share on average more DNA identically by descent than do individuals of Asian ancestry or of European ancestry broadly defined. Extensive population level sharing of identical genomic segments complicates the prediction of relationship level for a pair of individuals since identical segments may reflect common ancestry older than the most recent ancestor for a pair of individuals in a pedigree. In order to understand the pattern of the observed population-level sharing, we simulated extended pedigrees using different populations to calculate the expected amounts of sharing for 1st through 10th cousins. Specifically we assessed the relationship between the length of the longest segment and the most recent common ancestor. From the simulations, we also determined bounds for predicted cousinships given a specific amount of segmental sharing. Using these bounds as a guide, we detected at least one distant relative, between 2nd-8th degree cousin, for 90% of individuals in our dataset of more than 5000 individuals with European ancestry. An even higher fraction of our Ashkenazi sample, 99%, had at least one distant relative between 2nd-7th degree cousinship in our dataset.”