Ancient DNA Calculator Update

Felix Chandrakumar’s separate Neanderthal, Denisovan and Clovis Anzick tools have been retired in favor of a combined calculator that will compare your DNA to 31 archaic and ancient samples.  It is 131 MB and can be freely downloaded here:

Featured image

Felix explains, “The tool checks for small segments chunks within the DNA to see how much they are identical. You need to remove the genealogical perspective and put on the general idea that every human is 99.7% identical. This tool tells you how much of that 0.3 is identical. Instead of SNP by SNP, which mostly match each other by 60 to 70%, it checks smaller chunks of identical segments.

The 99.7% is based on each SNP comparison. If we do the same, everyone will match at ~60%. The only problem is, that does not tell the true shared DNA but shared SNPs. If we compare smaller segments as in genetic genealogy like 1 cM some will never match and that’s not true either. However, the goal here is to find shared DNA which is a thin blurred line between both. So, the idea here is to take identical chunks of small segments greater than 100k positions also considering for runs of Homozygosity, the percentage of matching is calculated. The reason why 100k is considered is because most gene lengths fall within this length range and the noise/accidentaloccurrence range of 1-2% is acceptable.”

I ran the raw data of a few family members through the archaic tools and came up with these results:

Featured image

More can be found out about each individual ancient sample and its age here.  All of the samples in the tool are uploaded to Gedmatch and are available to run your kit against with the one to one utility.

Ongoing discussion of the Ancient and Archaic tools at FTDNA forums can be found here.

DNA Calculator: Another reason to IGNORE Small Segments

23/12/14 edit to add: From Felix Chandrakumar, ” I just updated the DNA calculator to customize thresholds/settings along with 2 predefined options for total and recent ancestry. Hope this will avoid confusion. I also made the other 3 tools obsolete.”

You will no longer be  able to replicate my experiment, but you will still be able to calculate your total recent ancestry.  Enjoy!

Felix Chandrakumar has come out with another gadget to amuse the DNA aficionados over the holiday season called the DNA Calculator.    According to Felix, “DNA Calculator will accurately tell you how much percentage of DNA is shared between two people. The accuracy is within the range of 1-2%. It supports FTDNA, 23andMe and Ancestry files.”

As an administrator for a few family members, I was able to run our raw data through the DNA Calculator and determine our shared percentages:

Featured image

The results were odd for several reasons.  To begin with,the percentage of DNA shared between close relatives was not what one expects to see.  ISOGG has published the following statistics:

  • Grandparent and Grandchild on average share 25% DNA
  • Great-grandparent and great grandchild on average share 12.5% DNA

Why was I sharing 39-45% of my DNA with my grandmothers?  Why were my sons sharing between 35-41% of their DNA with their great grandmothers?

In addition to the seemingly weird shared percentages, the entries marked with asterisks aren’t related in a genealogical time frame.  They don’t appear on each others match lists at FTDNA or Gedmatch, nor do they come from similar geographical areas.

  • Maternal Grandmother and Paternal Grandmother sharing 31.67%

I was surprised enough by the high percentage of shared DNA among unrelated people to not only re-run their files but to also re-download the raw data and make certain it had not been mis-labeled.

At Gedmatch I compared my grandmothers head to head:

Comparing Kit F364959 (*Duffy) and F372234 (*Kern)
Minimum threshold size to be included in total = 700 SNPs
Mismatch-bunching Limit = 350 SNPs
Minimum segment cM to be included in total = 7.0 cM
Largest segment = 0.0 cM
Total of segments > 7 cM = 0.0 cM
(1813) No shared DNA segments found

When I dropped their default values at Gedmatch to 100 snps/1 cm I began to get an idea as to why they had such strange shared percentages.  Across all of their chromosomes, the results were similar to the figures below:

Featured imageFeatured image

A person’s total amount of DNA is 6766.2 cm.  Gedmatch showed my grandmothers sharing 1253.3 cm, with the vast majority of their shared segments  in the 1-2 cm range, and a handful up to 4.8 cm. With my grandmothers sharing about 16% of their DNA at mostly 1-2cm segments, the rest of their shared DNA is beneath the 1 cm  default I had set.

22/12/2014 edit to add:  Felix Chandrakumar confirmed the conclusion I was coming to, “If I want to be clear, the tool is meant to know the shared DNA which tells your total percentage of shared ancestors in each others pedigree.”  None of these ancestors are recent.  The shared ancestry isn’t even population based.  It is human ancestry.  We all descend from the same original population and are of course going to have many shared ancestors in the human pedigree.

The DNA Calculator. is a fun enough bit of kit for immediate relatives or even unrelated individuals, as long as the user is aware that all DNA is being measured, not just segments above 7 cm/700 snps and the total shared DNA % will reflect that.  Given the ongoing debate of how low you can go in setting default values when comparing kits, this tool is a useful illustration that those very small segment matches are in everyone’s DNA, not just your matches.

New DNA Toys: Archaic Human Percentages

23/12/14 edit to add: From Felix Chandrakumar, ” I just updated the DNA calculator to customize thresholds/settings along with 2 predefined options for total and recent ancestry. Hope this will avoid confusion. I also made the other 3 tools obsolete.”  Hopefully, Felix will change his mind and reintroduce the tools.  

23/12/14 edit to add: Felix has updated the tools and made them available for download here.  You can see the results of my experiment here.

Just in time for the holidays, Felix Chandrakumar has given us some new freebie genetic genealogy tools to play with. One of my very few disappointments in testing through FTDNA is the lack of archaic DNA percentages given to participants.  With Felix’s new tools and your raw data, you can compare your DNA to archaic human samples.   The Neanderthal Tool can be found here and the Denisova tool here.  He has also made a tool to compare yourself to the 12,500 year old Clovis Anzick sample found in Montana, which can be found here. Featured image I found the percentages astoundingly high until I read Felix’s clarification, “The tool checks for small segments chunks within the DNA to see how much they are identical. You need to remove the genealogical perspective and put on the general idea that every human is 99.7% identical. This tool tells you how much of that 0.3 is identical.” National Geographic provides answers to how we share a common genetic heritage with the archaic humans Denisovans here and about Neanderthals here.  Our results still seem a bit high to me compared to the numbers given by 23 and Me and National Geographic testers.  Have you tried Felix’s tools yet?   What do you think?

Ancient DNA Matching

Thanks to Felix Chandrakumar, who has processed publicly available Ancient DNA and uploaded it to GEDMatch, it is possible for anyone who has a kit at GEDMatch to compare themselves to DNA dating back 50,000 years ago. These ancient samples have created quite a stir in the genetic genealogy community as when the results became available, not only did it become clear that the ancient samples were matching modern testers, they were matching them in what appeared to be a genealogical time frame.

Obviously, no one who lived 12 thousand years ago is going to have living relatives with a common ancestor is the last 5 generations.  So, what on earth was going on?  Sample contamination was considered, as was the possibility that the age of the DNA samples themselves was incorrect, and they were in fact much younger.  Yet, with ancient sample after ancient sample, modern matches were appearing. Possibly, many of us simply carry similar DNA  that has become common to the make up of most humans on the planet.

As a rule of thumb, those who study genetic genealogy consider any match who shares blocks less than 7 centimorgans as being a 30-70 chance of being a true, or Identical By Descent match.  However, when studying ancient DNA, the 7 cm. cut off was ignored,  as living matches were not expected.  GEDMatch offered a tool to test individual kits against all ancient DNA samples, and allows the minimum centimorgan level to be set as low as .5 cm.

My kit matches many of the ancient samples at very low levels.  As seen below, each line represents a DNA block the length of 2 centimorgans that I share with each sample, and there is nothing even slightly unusual about that: Featured image What was unusual was that when I raised the centimorgan cut off, Loschbour and I continued to match on Chromosome 8: Featured image None of the other ancient samples produced anything like a similar result for me.  Upon further investigation, I discovered Loschbour had other modern matches, with the highest match coming in at 11.2 cm. So who was this ancient man and why was my DNA matching his?

I checked Loschbour  and discovered he was an, “8,000 year old skeleton from the Loschbour rock shelter in Heffingen, Luxembourg.” Luxembourg is an area I’m familiar with, so I was delighted. Both sides of my family have ties to the same general area.   As my dad has also tested and has a kit at Gedmatch, I was able to compare him. He did have a single 4 cm match to Loschbour, but had no match to any of the ancient samples beyond that point.  Most importantly, his 4 cm match wasn’t on Chromosome 8, but rather on Chromosome 2.   Meanwhile, my maternal line descends from the immigrant Margaretha Toemmes, born in 1815,  whose family lived in the village of Trassem (Trier) for more than 500 years.  Trassem is only 43 kilometers from Heffingen,where the skeleton was found. Featured image Families that have lived in Trassem for centuries, like any other small village in Europe have intermarried repeatedly and mine is no different.  I have had no problem confirming this family line back to my 4th great grandparents,  Johann Temmes and Maria Schmidt, as descendents essentially carry a ‘double dose’ of DNA thanks to cousin marriages and matches are easily spotted.  Given our relatively close cousin-ship with Loschbour, it appears those cousin marriages may have stretched much further back in time than we realized.

Mary Toemmes Lebold

Examining the picture of Mary Toemmes Lebold (the daughter of Johann and Maria Schmitt Temmes), we never thought her to look particularly Germanic and I’ve often wondered what the story was with the ancestor that gave her features such a distinctive cast.  Originally, when cousin Bernhard helped me dive into the Church Records, I wouldn’t have batted an eye to find a Hun or Tatar in the family tree, given Mary’s looks.  Instead, I found centuries worth of generations all carefully documented in the same little village.  Cousin Bernhard had no explanation for Mary’s appearance, either, other than that Germany had been a crossroads since ancient times.

So, what does ancient DNA have to do with Iowa or Iowan DNA?  More than you might guess, if your family is anything like mine and you have ancestors that spent generation after generation in the same locality, and then eventually immigrated to America. My Toemmes family arrived just in time to fight in the Civil War, first living in Marshall, Illinois and then moving on to Iowa, where my grandfather James Wise raised his family.  Given the number of ancient DNA samples available, with more being added,  you might also get a surprisingly substantial match that agrees with your paper trail.