Centimorgans or Percentages?

Average Estimated cMs for Autosomal Testing Comparisons. click to enlarge
Kristina Gow Dunnaway, ISOGG Facebook Page, 2014, used with permission

I will take a moment away from the narrative to answer the most commonly presented question put to me by readers regarding autosomal DNA matching.

And yes, I just heard that audible group sigh from all of you chomping at the bit to hear about my mother's reunion with her father. All in due time. All in due time. 

But let's start out the new year right. It is important for all genealogists - novice and experienced alike - to start 2015 with a cheek swab or a vial of saliva. And if your response is that you have already done so, then you need to start 2015 getting your older relatives, who are regrettably finite resources, to spit or scrape. Remember that DNA testing benefits both you, the researcher, and those out there desperately looking for a match. Take a lesson from what you have read in my blog. My path to Harold James "Brighton" Daugherty, my grandfather, began with my mother's relatively tiny 0.52% match to Brighton's first cousin, thrice removed: Brian Joseph Ryder. 

Brian's great-great-grandmother, Bertha Daugherty, was a woman who died at the age of thirty-six years: sixty-three years before Brian Ryder was born. He never knew the woman. He never knew her name. When he started to poke around into his ancestry at the time of his 23andMe test, she became a passing fill-in-the-blank on his family tree. But Bertha's brother, Ira Daugherty, was my great-grandfather, and Ira's son was the man I spent 2014 searching for.

If Brian Ryder had not tested out of sheer curiosity, I would not have had the starting point for my search.

The larger the various DNA databases become, the more helpful they will become to the genealogist, the adoptee searching for his or her birthparents, the foundling with no history at all, the millions of children born of sperm and egg donations that have made modern-day genealogy so technologically baffling. DNA testing helps everyone.

Do it. Do it now.

So that takes me back to the question I am asked most frequently.

When I discuss relationships and DNA matching, I often do so in terms of percentages. It is one way that 23andMe lists their genetic matches, and the mathematics makes more sense to me and my analytical brain. I have posted a graphic with my blogs indicating how known relationships should theoretically match each other by percentages. Siblings match each other by 50%. Half-siblings match each other by 25%. First cousins match each other by 12.5%. And the biggest revelation for my search came when my mother matched Ken Ryder by over 4%; and I knew second cousins match on the average of 3.125%.

But not all DNA sites list percentages. And the total amount of DNA tested by each company varies slightly, as well as how they report it. Additionally, the percentages by which different sexes match is skewed a bit by counting the matches on the X-chromosome, as women have two of these to the man's one. Roughly, the centimorgans of DNA you match with another person divided by 6800-7100 should give you a ballpark percentage.

What the hell is a centimorgan anyway?

Wikipedia defines it this way. "In genetics, a centimorgan (abbreviated cM) ... is a unit for measuring genetic linkage. It is defined as the distance between chromosome positions (also termed, loci or markers) for which the expected average number of intervening chromosomal crossovers in a single generation is 0.01. It is often used to infer distance along a chromosome. It is not a true physical distance however."

Confused? Don't be. What I wanted to point out by this definition is that a centimorgan is not a tangible distance, such as an inch or a centimeter. It does infer a length of segmentation along a chromosome, and as genealogists we can think of it as a "sort of distance." All DNA testing companies report the amount of DNA you share with a match in centimorgans.  When you upload your results to GEDmatch.com, the many user tools also show matches in centimorgans and not percentages.

Oh, and when I say you should all get autosomal DNA tested in 2015, that automatically means you have to follow up with an upload of your results to GEDmatch.com. If you are only going to test through one company (cheapskate), you can still compare your results to others on GEDmatch.com who have tested through other platforms, and who have also uploaded their results to this site. It's free. Free is good.

Do it. Do it now.

So I am presenting you here with a chart similar to the one I have posted before in which the percentages of DNA are shown that you have in common with known stated relationships. This chart I give you today shows you the theoretical average of shared DNA you have with known stated relationships in centimorgans. This handy chart was made by Kristina Gow Dunnaway, and she gives permission for its reproduction and personal use. If you publish a book with this chart included and make a ton of money, that's another story, but I will leave copyright law to Judy Russell at Home - The Legal Genealogist.

You will see that the chart uses 6800 cM of autosomal DNA (atDNA) as its base figure for total DNA measured per person. This is the amount tested by FamilyTreeDNA. A more detailed discussion regarding the numbers game, the testing companies, and counting the pesky X-chromosome can be found at the International Society of Genetic Genealogy (ISOGG)'s wiki page at Autosomal DNA statistics - ISOGG Wiki. I have visited this page so often my browser recognizes it as soon as I type "au" only.

And remember, Mother Nature does not follow the rules set out on either one of the charts that I have given you. These are averages. The numbers are based on a purely theoretical assumption that DNA is passed perpetually in a tidy 50:50 split every generation. It is not. The only true 50:50 split you will ever get is a child compared to his or her parents.

The key to remember is that the larger the number, the more reliable the relationship assessment should be. I knew at the beginning of my search that my mother's father was not the man she thought he was, because she matched her sister by only 26% (1935 cM). There is no way you can make an argument for that being a full-sibling relationship. But as the numbers become smaller and smaller, the known relationship gets fuzzier and fuzzier.

Additionally, remember that if you have cousin marriages in your ancestry or come from a highly admixed population that may have had limited choices for marital partners, due to say religion or perhaps geographical isolation, the numbers become wonkier and less defining. The more families intermarry and their common ancestors' DNA is "reinserted" into their offspring, the more of it will be passed to the present generation. The numbers will be larger than expected.

Remember when I said life was messy?

Get busy setting up the 2015 budget, and make sure there are ample resources set aside for DNA!