Proving Ancestors through "Chromosome Mapping".

Wanda, there's a problem with this methodology.

In order to establish that a particular DNA segment came from a particular person, you have to know all the ancestors of a the people involved in a particular generation. If you don't, you can't do the triangulation correctly.

There is an easy way to picture the problem. If two people have a matching DNA segment and you want to find the common ancestor they got it from, you have to know all their ancestors in that time period in order to find the one ancestor (or ancestral couple) who can be the only answer. Otherwise, they could have many different ancestors in common.

And that means this sort of analysis -- that Tudor is an ancestor of Welles -- is a DNA theory not a DNA fact. And, it's not a particularly useful theory because there would be literally thousands of unknown ancestors that could account for the same data.

This is true Justin. Welles is fairly easy to establish and Tudor is only a possibility as the surname Tudor comes up as a match for the end location number in question, but more people with the Tudor surname would need to be part of the research.

The descendants of Halfdan are numerous. The end location number for Clifford, Bourchier and Dutton is the same, meaning they all come from the same line. They are also related to each other through the same line. Anyone with these ancestors: Henry Beaucler I King of England, Rollo, et all, can find the evidence of that DNA segment on chromosome 11, End location number 56273717. Ref: http://chromosomemappingofancientbloodlines.blogspot.com/2017/05/th...

Wanda, identifying matches by end location is an interesting strategy, but still you it's hard to see how you can ever overcome the problem of ruling out other lines of descent when someone has unknown ancestors.

Thank you for responding Justin. We will always have unknown ancestors. Especially, when we go past the point of known records. But thanks to all the hard work of curators like yourself and others, we know more now, then we have ever known.

Some "End Location Numbers" are numbers associated with many different surnames. This is due to the adoption of surnames by people who were not always biologically related. Then there are Irish or Scottish Sept's of Men who adopted the surname of their leader. There is the issue of patronymic naming system where the surname may change every few generations in the fashion of Scandinavians and the Welsh. Finally there is the issue of non paternity events. Chromosome Mapping by End Location Number relevant to Surname can help separate these variables, but it does take group participation and collaboration when there is doubt or when too many variables exist.

We do run in to unknown's that show a pattern. Despite the question mark, we add them to the database anyway because sooner or later, someone will have the missing ancestor or line of ancestors that matches that end location number. Every pattern found is useful but we don't add useless information to the database. There is always at least one of three matches who have a tree listing the ancestor with DOB/DOD and who they married.

The way the method works is by using a computer algorithm pattern. When you enter a surname in the ancestral search bar on family finder, familytreedna.com, the results will show you all the people who listed that surname in their family tree. Not all of them will match and not all of them will have inherited the DNA segment of that ancestor. Some of them will be invisible matches, some will be unknowns, and some will be spot on. By looking for and finding a pattern of at least three matching end location numbers through the chromosome tables for a specific surname means you have found an inherited DNA segment relevant to that surname. The next step is to look in your matches tree so see who their ancestor is. If their ancestor shares the same line as yours, then you have found the IBD, YDNA line for that surname. That's the very best you can hope for. You may also find patterns for the same surname, but not know who or how they are related to you. It's still useful information as it will be useful to someone else who does have that line and matches it. We have thousands of ancestors and don't always know who they all are.

When DNA companies offered us the prospect of working on our family tree's by connecting us with others who share DNA with us down to as little as 1cM, they did not tell us who we matched specifically in regard to DNA segments. In other words, all the suspected matches are lumped together and we are presumed to match by DNA generally with someone by some ancestor). That is the method that genetic genealogists and scientists use. It's a nice easy out; because it's safe to say you share DNA with someone when it shows up in your gedmatch, but other than that, they have not been able to say specifically as to who, how much and who passed it down.

I wondered why (in the beginning) that I did not get specific matching segments on the chromosome table to every person I matched by surname, so I printed out every single match for a surname and started looking for a pattern. That is when I discovered that there indeed was a pattern. (Discovery #1) The pattern was the "End Location Number". I found that the Surname had a pattern to the End Location Number of the Centimorgan on the chromosome table. I also discovered (Discovery #2) that the Start Location Number was relevant to the distance to the common ancestor and that the "Start Location Number" can change with the Surname in question, but the "End Location Number" stays the same down to every single digit,(unless a mutation has occurred). The third discovery (Discovery #3) was that some "End Location Number's" were found to connect to Male YDNA Haplogroup's. The fifth finding (Discovery #5) is very important because it goes against everything we have been taught about DNA inheritance thus far. We have been told that we inherit 50 percent of our DNA from our Father and 50 percent from our Mother. In general terms this is true, but in specificity, it is not true at all. By mapping out my Father's line, I found on average that his Surname matches amounted to an average of 7-9 cM on average. This is no where close to 50%. In fact, using the family finder Centimorgan table, it's a tiny portion of the 1,580.24 total Centimorgans of DNA including the X chromosome. This means the other 1,579 or so of Centormorgans of DNA is coming from other ancestors, including the maternal line.

Finding six (Discovery # 6) is that 1cM is not noise at all. Many relevant lines have been found showing a solid pattern for a Surname that includes 1cM matches. True, they are quite ancient. How far back we can trace them, remains to be seen but again, the cM amount and the start location number relevant to the cM signifies distance to the common ancestor. We have found for the same surname matches that are as small as 1cM in length up to 10cM in length coming from the same line. Even though it is 1cM, it did not change or become noise because it is small, it merely represents an ancient common ancestor. These small cM matches can remain constant over centuries of time. The reason why 1cM is not noise is because these small cM segments have been traced to surnames which have been traced to an IBD line coming from a YDNA line and a Haplogroup. We already know that Haplogroup's are not noise, so in relevance, neither are 1cM matches. People just didn't understand them or thought that their size was insignificant. In the end, they may end up being the most interesting segments of all.

In essence, what you are looking at when you see these stacks of colors on the chromosome tables in relevance to a specific surname; is an "Identical by Descent" male YDNA line. Some bloggers call them "Pile Up's" which doesn't sound very scientific, but in reality, they just don't understand that some of those "Pile Up's" are IBD male YDNA lines", but ONLY if they line up on the right and every single digit matches for the end location number to the Surname.

This will make sense for people who use family finder at familytreedna.com as many of our group members have been able to map out surnames in their tree using this method. It helps to piece together missing pieces, question marks and either confirm or not confirm if a person with the same surname is biologically related to the same line or different one. We have mapped out hundreds of surnames using this method so far.

Geni has been a huge help in researching those surnames to see if the end location number found, matches the pedigree on geni. In many instances is has and does.

I understand that everything I am saying goes against everything known and published by scientists and genealogists today and that I am no where close to being either. I don't pretend to be nor is it my educational mission. I am an "Entrepreneur and an Inventor". Entrepreneur's Innovate. We look at problems and search for solutions. Those solutions are sometimes Innovations or Inventions. I saw a problem with the matching system for DNA so I took it apart and looked for and found the pattern, studied the variables (Random DNA inheritance) then put together a system (a database) for collecting results using a method based on finding the DNA pattern of Surnames by mapping the "End Location Numbers", Surnames and ancestors of people sharing that Surname and End Location Number. I have been teaching the method to other people and they have found that everything I said is relevant and they have been able to map out their ancestors using this method. '

People can try it. Free instructions are posted online. Anyone is welcome to kick tires, prove or disprove it. For the most part we are pretty happy with it and have been pretty busing mapping out surnames. I hope that data will be useful in the future, with high hopes that this method will help us map out the human race and our connections to each other through our thousands of ancestors.

Thank you so much for your open mind and kind consideration. Hope this helps.

Again, any Curator on geni is invited to join the project if interested.

Wanda, it sounds very logical, doesn't it? The problem is that it doesn't really work. And it can't work because of a simple logical flaw.

The underlying assumption is that two or three people who share the same segment (with the same ending position) have a common ancestor who can be identified. That would mean the segment can be assigned to that ancestor.

Except it doesn't work that way. If I share a segment with someone, and it's long enough, we probably do share a common ancestor somewhere in our trees, but if either of us have unknown ancestors, we can't tell whether we got our shared DNA from that unknown ancestor.

I've worked through it again and again, trying to test your theory, and it always turns out that the common ancestor I share with this person or that person could be somewhere much more recent. Then, when I've found more info, it does in fact turn out that there's a much more recent connection that probably accounts for the ancestor.

There will always be variables and unknowns where autosomal DNA is concerned. We can only be certain of what we know "right now"; for example the DNA segment found on chromosome 11, end location number 56263717 matches Dutton to Dutton, Clifford to Clifford and Bourchier to Bourchier and Clifford, Dutton and Bourchier all have the same male ancestral line so it is logical to assume that DNA segment is being passed down through this Male YDNA line. cM and SNPS can help point to a time frame and the collaborative research and sharing of data so far has shown an average range of 1.30 cM and 600 SNPs for common ancestor Idonea de Clifford born circa 1303. All three matches for Idonea de Clifford as a common ancestor are EXACTLY 1.30 cM and 600 SNPS. Isaac Clifford born 1664 matches the SAME EXACT END LOCATION NUMBER, but the cM is 1.54 and 800 SNPs. This is reasonable and logical because it follows the information we have been taught that the closer the ancestor was born, the larger the cM and SNPs. Notice it did not change a lot? The reason why is because the common ancestor John de Clifford 7th Baron Clifford born in 1389, is only a half century in time but the cM did change from 1.30 to 1.54 cM which is very useful information. The SNPs changed from 600 SNPs to 800 SNPs which also gives us some very useful information.

The reason this method does work is because of the way the computer algorithm works. The familytreedna family finder program is designed to match people with other people who list a surname. We still have to research the listed ancestor in our matches tree, but geni is a huge help in that regard. Finding the end location pattern = surname - variables = a pattern for an end location number relevant to a surname. The possibility of the end location number relevant to a surname coming from a different ancestor is possible if that unknown ancestor passed it down in a specific line. It is only with data input from more people that we can reduce those possible variables. However, we can still attribute a surname to an end location number by doing the research and finding the pattern and know that "right now" we know that end location number is coming from X, until proven otherwise. End location numbers have been found to be associated with haplogroups which means more than one surname can be attributed to a haplogroup. This method despite possible variables, is still better than what we have now. It works, it is repeatable and people are finding it to be useful.

In regard to the common ancestor you found that "could be" more recent, does everything add up? Did you find a common ancestor for the surname you are researching? That can help weed out a lot of variables. The other big variable is coming across a cluster of ancestors and trying to separate them by end location number. When we come across those, we add them to the database and wait till another person with a separate line is able to define which ancestor the end location number belongs to. I came across that myself with an entire family who matched me on about 8 chromosomes and as many surnames very closely. You can't make a mistake by entering the data in the database as you find it, even if you are not sure right now. In time, through collaboration, it will sort it's self out. An example is if you have a group of Grandparents named Smith, Jones, Rice and Henry with another family that also has those ancestors born around the same time and place. It would take a third person who only has one of those ancestors in each case to separate them out. It might take a little longer but collaboration really is the key.

I think you have not read or understood what I'm saying.

This type of analysis works very well in recent generations, where the odds are greatly against a DNA relationship that does not actually follow the surnames two people have in common.

However, the number of ancestors expands exponentially each generation. That means the identity of all the unknown ancestors becomes a barrier to mapping the DNA to any one of them.

In other words, if you have a Clifford ancestor and I have a Clifford ancestor, both in the 1500s, and we have a DNA match, it's much more likely our match is through one of our more recent unknown ancestors than it is that our match is through the Cliffords.

Family Finder at FamilyTreeDNA.com allows you to search by surname. You will only be matched with people who have listed that surname in their list of ancestors. Not all of them will inherit the DNA segment. Some will have incorrect or broken trees and a large majority have not tree's at all. When you do find matches, you have to look at all the sets of matches and filter out other possible ancestors. The best case scenario is when you find the "End Location Number" Pattern that matches the Surname AND the Pedigree matches as well. In other words, you share a common ancestor in that same line for that same surname and you share that surname with more than one person and everyone has the same Surname, the same Pedigree, Common ancestors in the same line and the End Location Number on the same chromosome is the same as well. Our research has shown that some of these matches include IBD/YDNA comparing to Matching by Markers only better. The matching by markers through YDNA had group together Men who are related but they may only be related by a group of people born anywhere from a few hundred to a few thousand years ago. Once you go past a thousand years or so, surnames don't really apply. so autosomal DNA really works better in finding ancestors that we can actually trace. Not knowing the name of an ancestor born 2000 yrs. ago is not a barrier as we likely would not know their name anyway. The Clifford line that we have mapped out is confirmed by more than one person. It is a true pattern. That line shows up in just about everyone who has on of those ancestors in that line. You can spot it by the amount of cM and SNPs. You might think it is a recent ancestor because it shows up so often and in so many matches but it is not until you get a larger matching cM and SNP that the possibility can exist for this being a more recent ancestor. We have between 120-130 entries in the database so far. A lot of entries are made for descendants of this line because it is so easy to find, but again, it is the cM and SNPs that will tell you how far away that common ancestor is.

Wanda, you are pushing the DNA far beyond what it can really show.

Thank you :)

Here are two very good videos that explain the limitations of using DNA for lines this far back.

Are You Genetically Related to Charlemagne? (4:33)
https://youtu.be/_vm6almBFEo

Why am I only genetically related to 120 ancestors? (8:33)
https://youtu.be/9g-VwtRd8ks

Thank you Justin. Those are great links. In the first reference, we are trying to map out the "End Location Number" for Charlemagne. So far, a specific end location number on Chromosome 5 looks promising. It also appears to be a SUPER NUMBER, which is what you would see with anyone who had an enormous amount of descendants.

The end location number on chromosome 11, associated with descendants of Fornjot (some call him mythical but we use him as a pin to see where his last male descendant is so we can track the pattern of inheritance. The last male descendant from Fornjot is the IBD line and those following are autosomal. All of his descendants who had children with the cousins of his other descendants have made it possible to pass that DNA segment down and around enough that just about anyone can find it in their matches. Anyone with a tree on geni going back far enough can pin Fornjot and see who the last male in their line from Fornjot is and get an idea who may have passed it down as well as the common ancestor. Chromosome mapping by end location number works with autosomal segments which are far more numerous but it also works with finding nice long segments of IBD lines by using geni as a research tool.

So far, in our research project, we have found two SUPER NUMBERS: The Kings Line descending from the Ancient Kings of Kvenland (if you prefer, pick one of his male descendants if you don't like the idea of Fornjot. That number is End Location Number 56273717 found on chromosome 11.

The other SUPER NUMBER found is associated with Niall of the Nine Hostages, End location number 137439496 found on chromosome 2.

Finding the End Location Number for Charlemagne is harder, even though everyone is related to him. The problem is firstly, finding his last male descendants in his line and having a surname to work with. So far, it has been narrowed down to the Counts of Vermandois, Louvain or Fitzherbert. Louvain is tough because no one puts Louvain in their list of surnames on family finder. However, they do list Herbert and Fitzherbert but there is a bit of a conundrum there. I did find two different Herbert lines. Both are on chromosome 5. They are different End Location Numbers meaning different lines. The first set of Herbert line matches are much smaller in number as far as matches go. I did get one match to Heribert and the rest to other Herbert's, but it is nothing like what you would expect from a SUPER NUMBER coming from someone like Charlemagne whom everyone is supposedly related to. While we can easily find our connection to him on geni, he has not had any male descendants in a very long time. Finding his DNA being passed down after all this time would require his descendants to have been "busy" in the manner that Niall of the Nine Hostages was busy. He did have a lot of children without doubt but when you look at those children, not many of them had children. The other problem was him not wanting his Daughters to marry and have children. They were sent off to convents.

The other possible End Location Number found for Herbert matches is a number that could be considered a SUPER NUMBER. While the Count of Louvain "almost" has a direct connection to Louvain, there is one female in the line of descent from Charlemagne. That creates a variable that has to be considered. The Vermandois Count line runs into a surname problem so you have to look for a direct male descendant in that line with a surname you can actually research.

This website says that Herbert is the longest running male line from Charlemagne. https://www.eupedia.com/europe/families_descending_from_charlemagne...

While we have an idea of what End Location Number "might" be associated with Charlemagne, I don't think we have enough data to be certain. It is a long way back to his last male descendants. Finding surnames to work with is another problem but I think in time with enough data and people collaborating we can figure it out.

In reference to the second youtube video and us only having 120 ancestors, that video was made with "the information they had then". Things change. We learn and make new discoveries all the time. From the research we have done so far, it has been found that we inherit DNA from a lot more than 120 ancestors and we can even find it. To put this in scientific terms that make sense for everyone, please look at the chart https://www.slideshare.net/AncientBloodlines/chromosome-mapping-of-... of Centimogans of DNA we all have on each chromosome. Many of our matches are in the 1-2 cM range. If only half of all our matches are 1cM, that would equate to roughly 1500 common ancestors. If the rest of matches average 3cM, that would give us another 500 common ancestors and this is just an example of variables that are possible. Another fact to consider is; if a person has 5000 matches on family finder to cousins who have an average of 5 matching end location numbers on a chromosome table that would = 5000 x 5 = 50,000 possible ancestors in common. Of course, some will be duplicates and some will be biologically related because they share the same haplogroup but a lot of those matches will be unique. Can those 50,000 possible common ancestors be reduced to 120 ancestors? It's too soon to tell but when I run out surnames to map out and do a final count I'll let you know. This may take a lifetime though, so I hope you intend to A. Live a long time and B. Be very patient. lol

I haven't actually researched how many matches total fall in the 3cM or less range but they would be ancient common ancestors. Through our research, we are also finding that the these end location numbers are associated with hapolgroups and downline SNIPS from haplogroups. There are 49 Haplogroups. You will probably find a Smith on every one of them. In addition to all these Haplogroups, there are branches of each haplogroup. The H1 haplgroup just got an update with 200 new branches discovered. These branches help narrow down with more specification a unique line. https://blog.nationalgeographic.org/2013/12/24/how-rare-am-i-genogr... If there are 100 branches on 49 Haplogroups, then the possibility of finding a unique ancestor just increased exponentially, as well as the possibility of having a lot more than 120 ancestors. I've been mapping out my ancestors by these end location numbers for almost 2 yrs. now, and I have not run out of surnames to research or run out of unique numbers yet. If I only end up with 120 ancestors, I'll let you know, but it's starting to look like a lot more than that. :)

Wanda, if you listen to the videos and understand what they are saying, you will see right away that end location numbers -- essentially the breaking points for recombination -- are very helpful for recent ancestors, but not for more distant ancestors.

And a match at any level below about 7 cM is likely just background noise for the population as a whole.

Thank you Justin :) Oddly enough, we have found these End Location Numbers to be relevant to both close and distant ancestors in the same line. For example, End Location Number 1234567 on Chromosome 0 can range anywhere from 1cM to 10cM or more depending on when the common ancestor was born. Some of these go back about 1000 yrs. in time. The reason why the End Location Number does not change in that length of time is because a mutation has not occurred. The end location number is relevant to a haplogroup or a branch on a haplogroup. It can take a long time for those mutations to occur. These autosomal matches are based on IBD lines which might be 5 generations or 30 generations depending on how long that line existed without mutation. It's very interesting to find and study. The reason why 1cM is not noise is because it is based on a haplogroup, branch and surname, which is what we are finding using this method.

This discussion reminds me of a fun little book called Change Your DNA, Change Your Life, by Dr. Robert V. Gerard (2006). It's all about how activating your DNA with specific mental instructions can "improve your physical, emotional, and social well-being".

Not my kind of "science" but some people like it.

I wonder if we can also change our ancestors by changing our DNA. I'll see if I can find someone to ask.

That sounds like a very interesting book Justin. I saw online a discussion where a fellow said that mutations can happen when great stress occurs such as battle, plague and things of that nature. I had always thought mutations occurred with too many close cousin marriages. I don't know for sure, but perhaps the DNA mutation occurs spontaneously to create diversity in the genes so that in case another cousin marriage happens down the line that recombination is not occurring with the exact same set of DNA segments. The mutation of course is not always a positive thing.

Here's an interesting blog post by expert Blaine Bettinger:

Why Small DNA Segments are like poison M&Ms
https://thegeneticgenealogist.com/2017/12/29/a-small-segment-round-up/

Thanks Justin. I've seen that. :)

It seems to be time to remind people about the limitations of DNA evidence, and particularly the problem of short segments and confirmation bias. It's our modern equivalent of the genealogical frauds of past eras.

Here is a link to a presentation at RootsTech 2019:
https://youtu.be/UVIEp6O99iw?t=8947

Great lecture Justin, and very responsible. I like that the lecturer considers both DNA and Records and also looks for Mis-atributed Parents. Thank you for sharing. In response to "noise", I feel that people are forgetting that autosomal DNA is made of both MTDNA and YDNA and both of those kinds of DNA have been traced back thousands if not tens of thousands of years. Taking into consideration, mutations and subclades of haplgrops that happen over long periods of time, it is not reasonable to think that all our our autosomal DNA becomes noise after so many generations. If that was the case, then we descend from nothing and no one. I believe that long generations of Identical by descent lines of MTDNA and YDNA can stick around a lot longer than 4-5 generations. I believe the only autosomal DNA that becomes untraceable is that which is pushed out by stronger lines and repeat ancestors (those lines that have married back into your line or when any of our Parents or Grandparents become Cousins by a shared line. I simply don't believe that we are divided in half repeatedly going back each generation until we become zeros. It doesn't even make good sense scientifically. If MTDNA and YDNA can exist tens of thousands of years and be traceable and all of our Autosomal DNA is composed of that, then why would our Autosomal DNA disappear? This is what I believe, based on chromosome mapping. We have 22 chromosomes plus the X chromosome. I believe the space on each chromosome called cM and SNPs is like real estate. It can only hold so much information. Conventional Geneaology says that real estate is composed of your Mother and Father's DNA passed down to you from their four Grandparents and so on getting smaller and smaller with each generation. The problem is it is divided up into pie that gets smaller with each receding generation and does not take into account repeat lines, cousin marriages or long lines of male or female ancestors. It is those long lines that make it possible to map out ancient bloodlines. In other words, it is not a pie theory, but a line theory. For example,if an identical by descent male in your line is represented by + and female by - and your 2nd Great Grandfather's line goes like this: ++++++++++ -(10 generations of Swanstrom, then a female) and your 2nd Great Grandmother's line goes like this ---+-----++ (3 generations of females, 1 generation of male, 5 generations female, 2 generations of males), then the odds of your Father's line of ancestors being traceable is excellent (providing matches signed up with the same DNA company as you and share ancestors with you. Your Mother's line, because she is MTDNA, will probably be invisible for matches until someone in your Mother's line shares the same ancestor in your Father's line. If perchance your 13th ancestor through your Mother is a Swan, then your odds of finding matches born 13 generations ago are pretty good. It's not so simple as a pie. It's more like matching up long lines of morse code horizontally. It is not like a pie that gets divided into nothing, noise or zeros. You certainly don't descend from nothing and if you have a 30 great grandchild in the future, you certainly would not tell them you are just some noise from the past.

Hi Justin and Wanda Marie. Maybe you can help me with a simple question: Does Geni allow for chromosome mapping, the way FamilyTreeDNA, MyHeritage, and GEDmatch all do? To be clear, I'm talking about finding a cousin, and then determining which DNA segments we share - which chromosome, and what position on that chromosome. I've scoured the website and can't find that feature anywhere.

Thank you,
Doug

Doug Waggoner - No, but since Geni only shows you DNA matches to folks who are your matches at FTDNA and have also linked their FTDNA account and their Geni profile, look at FTDNA to see that info.

Thank you, Lois. I didn't know that FTdna is the only test provider linked to the Geni site. That makes things much easier.