Kristine and anyone else interested in genetic genealogy,

In one of your posts you stated that you wonder how accurate FTDNA matching is because you did not see common surnames. Katy had also stated that she would need to see your tree to determine if that match is correct or not. Even though the two of you have found common surnames I would like to expand on that misconception and how it can be cause people to erroneously shrug off matches and how to correctly determine if a match is real if you run into the same situation with other people. The common DNA isn't always from the common surnames and not finding common surnames doesn't always mean a match isn't real. All three DNA companies (AncestryDNA,FTDNA, and 23andme) use various algorithms to determine if two people are a match. There will be some false matches with all three companies so multiple things have to be done to determine if the match is real. Due to a misunderstanding of how the matching works and how DNA is passed on through the generations the misconception that no common surnames in trees must be due to inaccurate matching is one of the most common reasons people too easily shrug off matches.

Since the DNA that we inherit can be passed on for 15 generations then, in some cases, each of your matches have to have a very well documented tree for all of their ancestors going back as close to 15 generations as possible and without errors in order to find a common surnames. Many people don't know how, don't have time, or don't care to build an accurate tree that far back (this also happens at Ancestry.com where some trees only go back 5 or 6 generations and some others even less) or include all of the surnames in the surnames list at FTDNA. Just a few missing ancestral lines from just 4 or 5 generations or even as far back as 12 generations could be the cause of not finding a common surname. We also have to consider the small amount of affairs where the child doesn't genetically belong to the father. It didn't happen often but it did happen at times.

So what needs to be done?
Triangulation, as I had posted in in this thread, and several other threads on this site, and the use of DNA results of parent-child trios if possible. If parent-child trios aren't possible use DNA results of other close relatives. If a person matches you they should also match one of your parents on the same chromosome and segment. If not it is called IBS (Identical by State) or what some call a false positive.If the match exists in a parent on the same chromosome and segment it is called IBD (Identical by Descent) and this is normally a segment inherited from an ancestor in the past 15 generations.The segment should at least be 7cM in size (in the past I had used 5cM based on what some sites had recommended but recent literature has shown that 7cM or above is best due to a lower IBS rate). Not matching a close relative doesn't necessarily mean the match isn't real. It could mean that your close relative just didn't inherit that DNA segment. If your matches also match your close relatives and it is on the same chromosome and segment then the match is very likely to be real if the segment is large enough.

What is triangulation?
It is a process of finding multiple people that match on a specific segment on a specific chromosome and also match each other. They all have to match each other on the same segment and you have to match all of them. If two people don't match each other and they match you then they are related to you from different sides of the family. This happens because one person matches on one of the chromosomes in a pair and the other person matches on another chromosome in the pair even though it is the same segment. If they match each other and you then it is from a common ancestor between all of you. The more people you can find like that the easier it will be to find the most likely common ancestor all of you have. Having more close relatives test will get better results.

How is triangulation done?
FTDNA, Gedmatch, and Genome Mate. FTDNA has the chromosome browser built in so you can compare which chromosome and segment you match other people with but only five at a time. This is where the beauty of Genome Mate comes in. You can download that file and import it into Genome Mate so you can compare all of your FTDNA matches one chromosome at a time. You can also use Gedmatch to do a one-to-one compare on each match from the one-to-many match list that isn't already in your FTDNA file and paste it into Genome Mate. Then select a chromosome and look for multiple people that match on the same segment and then click on their name then email them to ask them if they match the other people and if so if they have a tree that they are willing to share, if it's not already at FTDNA. Do that for the other matches also. You can also find them in Gedmatch sometimes by using the one-to-many search then doing a one-to-many on them to see if they also match your other matches and do one-to-one on the other matches. Over time you will find triangulation groups.

How purchasing the $39 FTDNA transfer helps!
Since quite a few of the Gedmatch matches are in the FTDNA file you are saving a lot of time by purchasing the FTDNA transfer and getting the chromosome browser file and match list file. You should also be aware that not all of your matches at FTDNA have uploaded to Gedmatch and you are missing out on matches from FTDNA even after you have uploaded to Gedmatch if you don't purchase the FTDNA transfer.

Evidence of DNA being inherited from 15 generations ago
Jim Bartlett, a respected keynote speaker at DNA conferences, has made some excellent comments on a forum recently that "we sell ourselves, and genetic genealogy, short when we say atDNA is limited to 5th or 6th cousins. All of the evidence is pointing to the atDNA tests being valid out to 7th-10th cousins (and maybe more) - it's our own Trees that are the limiting factor" http://archiver.rootsweb.ancestry.com/th/read/GENEALOGY-DNA/2014-12/141…

He goes on to say he has sticky segments that beat the odds that passed down maybe 10-15 generations. http://archiver.rootsweb.ancestry.com/th/read/GENEALOGY-DNA/2014-12/141…

Ann Turner also pointed to a site from Coop Lab at UC Davis that has an illustration of sticky segments going back 11 generations. http://gcbias.org/2013/11/11/how-does-your-number-of-genetic-ancestors-… If you look at the 10th and 11th generation for a max contribution that is 1.5% per contributing ancestor which is close to a 10cM segment.

Inversely, not all people that match in a genealogical tree will match genetically
There are fourth and fifth cousins that do not show as matches because they have each inherited DNA from different ancestors even though they have the same ancestors. Some of the DNA gets lost or whittled down to a size so small it doesn't meet the threshold of a match at any company. For illustration purposes use those tree from Coop Lab. To see what a genetic tree is like see http://gcbias.files.wordpress.com/2013/11/family_tree.png and to see what a genetic tree is like see http://gcbias.files.wordpress.com/2013/11/family_tree_w_trans_2.png

Armando

Kristine,

If you have del Real and Quesada lines, we're cousins. I still need to submit my Family Finder test to FTDNA, so hopefully we'll see a match there.

I'd like to see how we connect, so I'll email you privately.

Rosalinda