Kinney, McKinney & Variations
Lenhart & Variations
DNA Project notes
GKBopp

MEANING OF MATCHES AND NEAR-MATCHES OF TEST RESULTS
See also GENETIC DISTANCE - FTDNA CALCULATION
See also FTDNA’s Interpreting Genetic Distance links:
http://www.familytreedna.com/gdrules_67.html
http://www.familytreedna.com/gdrules_37.html
 http://www.familytreedna.com/gdrules_25.html
 http://www.familytreedna.com/gdrules_12.html
Revised  28 September 2007

 

 

DNA TESTING DOES NOT REPLACE TRADITIONAL GENEALOGICAL RESEARCH.

If you have a paper trail to an established DNA line for your surname or variation, you already know who your common ancestor is. In that case, you don't care about the below probability table.

" . . . a theoretical relationship indicated by the DNA has no names and dates except those of the test subjects themselves. Everything else must be supplied by the documentary evidence . . . . What the DNA really provides is the encouragement to go looking for the documentation, or the opposite clue that there is no relationship after all." (John Chandler post to GENEALOGY-DNA-[email protected] )

If you match someone, but have no paper trail, the table may be helpful but the DNA test does not replace traditional research. For example, in the case of a perfect match of 25/25, the below probability chart indicates the time to the most recent ancestor could be anywhere from 0-37 generations. Participants with a close match must compare their genealogies and decide the next appropriate step to further their research. Test results can narrow the "field" by eliminating some lines completely. For example, in our project the participant had reached a brick wall. Evidence pointed to a well-known "Dutch" line, but for various reasons, a well-known "English" line was also a possibility. The participant was prepared to hire a professional genealogist but wanted to see if DNA might narrow the focus. The test results matched to another participant associated with the "Dutch" line and did not match the "English" line participants. The "Dutch" participants are now working together privately on their research.

____________________________________________________

 

 

* Probability that the time to the most recent common ancestor (TMRCA) is no longer than this number of generations. For example, in a 25/25 match the probability of generations to the MRCA is:
     7 generations or less - 50% [about 175 years]
    23 generations or less - 90%
    30 generations or less - 95%
    Estimated range of generations: 0 - 37
    Estimated year range: 0 - 925
** Based on 25 years for each generation

Source:
Time to Most Recent Common Ancestry (TMRCA) Calculator Using Genetic Marker Similarity Between Two Individuals
Bruce Walsh (Ph.D. Population Geneticist) TMRCA Calculator

See also:
Genetics, Vol. 158, 897-912, June 2001, Copyright © 2001
Estimating the Time to the Most Recent Common Ancestor for the Y chromosome or Mitochondrial DNA for a Pair of Individuals
Bruce Walsh
http://www.genetics.org/cgi/content/full/158/2/897


NOTE TO NON-MATHEMATICAL TYPES (e.g., the webmaster) - To people like us some things on the chart seem strange. For example, it looks to us as if there is closer ancestry on a 12/12 match than on a 24/25 at the 50% level. But this is a probability table and the 12/12 probability assumes there are no other markers to be tested (hence a 100% match), whereas 24/25 is only a 96% match. Thus, if you test only 12 markers (and have a 12/12 match), then upgrade to the 25 with a 24/25 result, the TMRCA "increases" at the 50% level. However, it decreases at other levels because of the mathematical form of the curve on less than 100% matches. In other words, this makes sense to a math whiz (who will understand the above link). The bottom line for the rest of us: for best results take the 25 marker test (not the 12) and use the above TMRCA probabilities.

See FTDNA’s Interpreting Genetic Distance links:
http://www.familytreedna.com/gdrules_67.html
http://www.familytreedna.com/gdrules_37.html
 http://www.familytreedna.com/gdrules_25.html
 http://www.familytreedna.com/gdrules_12.html
See also: http://blairgenealogy.com/dna/FTDNATiP.html

___________________________________________________________________

<> A single test tells you nothing; you need others for comparison. A typical study (testing of many participants) will identify group patterns. For examples, look at result charts from various projects; there are links at DNA Surname Projects.

<> If you did not have an "expected match" review Frequently Asked Questions and False Paternal Events (including section about Faulty Research).

<> It is desirable to have two or more distant male cousins tested for each known line to prove the line to a common male ancestor and establish the identification of that line (haplotype). Test results with close haplotypes (no more than two differences) indicate there is a high probability that participants share a common ancestor. However, the tests do not indicate how many generations ago you shared that ancestor. Your MRCA - Most Recent Common Ancestor - is a matter of statistical probability. For more information about this, see FTDNA's Understanding the MRCA (offsite) and the associated (and very technical) information by Bruce Walsh (Ph.D., Population Geneticist) at TMRCA Calculator (offsite).

<> Participants report frequent breakthroughs resulting from testing combined with traditional genealogy. However, the meaning of the test results without accompanying paper trails is complicated, as indicated in this quote (bold/underline added) from the Graves Family (offsite) site:

We are not yet completely sure what constitutes a match versus a non-match. One definition is that a match of test results indicates descent from a common ancestor, and a non-match indicates no common ancestor within recent historical time. However, all evaluation of the meaning of DNA test result comparisons involves probabilities, and never gives results with 100% certainty. In general, when two [25 marker] test results match each other on all markers, they are said to be descended from a relatively recent ancestor. When they differ by one mutation, they are still considered to be descended from a common ancestor, and when they differ by two mutations, they are probably descended from a common ancestor but there is more uncertainty. However, the conclusions and the degree of certainty depend on how many markers are tested, which markers are tested, and at which markers the mutation(s) occurred. As more markers are tested the confidence in the conclusion increases, and not all markers mutate at the same rate. (For example, according to technical papers studied by Patrick Guinness, mutations at DYS390 (locus 2), DYS385 (loci 5 & 6), DYS439 (locus 9), and DYS389 (loci 10 & 12) occur more often than at other loci.) What that means in practical terms is that if all the mutations are at the locations that mutate more rapidly, then the time/generations to the most recent common ancestor is less than is calculated when it is assumed that all loci mutate at the same rate.

The number of generations to the Most Recent Common Ancestor (MRCA), using the assumption that all loci mutate at the same rate, are shown in the [FTDNA] table. It should be realized that this table is based on statistical distributions, and an actual mutation can happen anywhere from one generation to infinity.

[See Understanding the MRCA at FTDNA]

 

<> See also, Blair's study on the 12-Marker vs. 25-Marker issue: Marker Analysis
     What the Test Results Tell You
     http://blairgenealogy.com/dna/markers.html

~~~~~~~~~~~~~~~

MUTATION RATES

This is a complicated and hotly debated topic. Unless you are a scientist and/or mathematician it's difficult to understand. And even then, there is disagreement about this. Here's what three recent authors say:

Trace Your Roots with DNA: Using Genetic Tests to Explore Your Family Tree
Megan Smolenyak, Ann Turner
http://honoringourancestors.com/books.html (and elsewhere)
"Mutation rates have been studied in hundreds of father/son pairs, in deep-rooting pedigrees with a few persons representing many generations, in shallow but broad pedigrees with many descendants of a few men, and even in thousands of sperm collected from a few men. The range of values is from zero (none detected in the sample) to 0.004, but the overall average when all data are pooled hovers around 0.002 (0.2 percent) per marker per generation. Another way of stating this mutation rate is one change in 500 generations for each marker."  p. 198-9

DNA and Family History
Chris Pomery
 http://www.dnaandfamilyhistory.com (and elsewhere)
"Calculations used in genetic genealogy are built at present around a number of numerical assumptions, the most important of which is that mutations occur in the average STR-type marker roughly 1 in every 330-500 generational transmissions. It's known that each marker mutates at a slightly different rate, but not how varied those rates are or the specific rates for each marker."   p. 150

Unlocking Your Genetic History (includes chapters on Y and mtDNA)
Thomas H. Shawker, M.D
https://www.ngsgenealogy.org/BookStore (and elsewhere)
National Genealogical Society - NGS Guide Series
"We know mutations occur. It they didn't, everyone would have the same DNA and the test would be useless. We assume that within the genealogical time period, one mutation can occur and not negate the relationship." p. 192-3

OTHER NOTES

<>There is an average of a 1/500 CHANCE in every generation that a mutation will occur on each marker. This way of saying it makes it clear that you could have two generations in a row that mutate, only the probability of that is really, really low: 1/(500*500) = 1/250000 or .000004. It's the same thing really as saying every 500 generations, but the emphasis is on the probability, which is more descriptive.

<> There is at least one reported case of siblings with a 24/25 match (a mutation occurred in one of the brothers) - don't have link but think it was the Wells project. However, here's a report of a two step mutation between two brothers:
https://lists.rootsweb.com/hyperkitty/th/read/GENEALOGY-DNA/2004-11/1101781625.

<> Some markers appear to mutate more frequently than others.

Fastest mutating markers according to FTDNA (Facts & Genes, January 31, 2003 Volume 2, Issue 1) 464, 449, 439, 385, and 458.
In May 2003 FTDNA began reporting that 458 and 459 also mutate at a faster rate than others. On the other hand, in the above remarks by Graves, "according to technical papers studied by Patrick Guinness, mutations at DYS390 (locus 2), DYS385 (loci 5 & 6), DYS439 (locus 9), and DYS389 (loci 10 & 12) occur more often than at other loci...."

<> See additional discussion at the RootsWeb DNA Discussion List by searching on "mutation rate," "mutation rates," "mutation rate assumptions in question," etc. For example, some report the variations in some large surname studies differ from the variations in a mixed random group.

<>Markers Chart at FTDNA shows how the TMRCA decreases after about the 23-marker threshold.

<> Match requires surname (or a least close locale) to indicate MRCA (can have perfect matches with random people) - see FTDNA's newsletter: September 26, 2002 Volume 1, Issue 3 - http://www.familytreedna.com/facts_genes.asp

<> See also Markers

 

Kinney, McKinney & Variations
Lenhart & Variations
DNA Project notes
GKBopp