DNA services - what's your experience? Recommendations?

[Allie]

I just got back my mother's DNA profile using the Ancestry.com service. I'm new to this and still figuring out how to work with the information. Hoping some of those who have submitted DNA will share their experiences and recommendations with everyone. What services did you use? What type of testing? What information did you learn, and was it worth it? Did it help with any specifically Civil War related genealogy?

My hope in having my mom's DNA submitted was to confirm my discoveries made using traditional genealogical methods about her paternal ancestry. My mother was adopted by her mother's second husband; she knew her biological father but didn't know much about his family, and now most of his family is gone. I also have an ancestor on her maternal side who is a complete mystery to me - I know his surname was Woodson, and that he died about 1850, but that's it. It was my hope that cousin matches might reveal where to look for more information on the Woodsons I'm related to.

I wasn't all that interested in the ethnicity reports. Although my dad has a couple of ethnic mysteries in his ancestry - he may or may not have had some recent Cherokee ancestry through a great-grandmother from Oklahoma, and has some ancestors who may have been Jewish - my mom is pretty straightforward: she's a white lady whose ancestors all pretty much came from Western Europe and the British Isles before the American Revolution.

The ethnicity results revealed no real surprises: 52% Western Europe, 33% UK, an unexpected 6% Iberian (according to Ancestry several varieties of Mediterranean people can also cause this result), little Irish, little Scandinavian, a tiny bit of Russian. So, on to the cousin matches.

I have my mother's tree completed for several generations with the exception of that one elusive Woodson. Back to about the American Revolution most of this information is pretty solidly confirmed by sources. Many of her lines are well-documented back to about the 1600's. With this information, Ancestry was able to find 95 cousin matches with a shared ancestor name on both trees on the first day. This info confirmed several lines I was unsure about, including one ancestor who was believed to have been a descendant of Pocahontas - matches with confirmed lines indicate he probably was. The information also indicated that Mary Chew Buckner, who received dower property from Larkin Chew, but was not listed in his will, was very likely his daughter and not adopted as had been suggested. So far so good.

I did not get any close cousin matches on the Woodsons, but I did get several matches which indicate that I should be looking at a group of Woodsons who moved to Cumberland county Virginia from Henrico county. I can surmise that I am descended from one Woodson who lived in the 1600s - all matches have him as a common ancestor. Not a lot of information, but more than I had before.

As for meeting new cousins and adding new relatives, so far, most of the matches that do not show a common name on both trees are Greek to me. In one case my mom has a high likelihood of a 3rd cousin, she has her tree completely filled out to the 6th generation, and we have nothing in common. Possibly an illegitimate child in there somewhere, or an error in one of our trees? Or a relative of whoever my Woodson's mom was?

Most of my matches fall into the "distant cousin" range. None are closer than 3rd cousins. Only one was a third cousin, several fourth, the rest 5-8.

The suggested relatives and ancestry's new thing, the DNA circles, have not provided useful information so far. There are a bunch of people who share DNA with me for no reason that either side can discern.

Supposedly it takes a while for the system to populate matches and I can expect more.

So far, I would say my experience has been that it was worth it, but only because I had gone about as far as I could using conventional means and needed a new toy. For a starting researcher I'm not sure how useful it would be.

There are several things I plan to do now, such as making a public tree for others to look at (my tree is private at the moment because I use it as a scratch pad and it has good, bad, and downright crazy stuff on it), and downloading my DNA information to upload to other services.

Can anyone offer helpful information about other services that will use the same DNA results? Has anyone tried different types of testing, and how did it compare?

 

[Dave Wilma]

No DNA experience here. All my family history research is analog and pre World Wide Web. I discovered a couple of surprises such as my wife's Irish grandmother in New York was Jewish. I suggest trying everything you can to build the family tree and use the DNA as a confirmation or as a means of inviting new avenues of inquiry.

 

[sonofboth]

You can't really discern British from Western Europe very well, particularly given the Anglo Saxon',s Jutes ,Danish Vikings etc. from northern Germany. Genetically the lowlands and most of the English are the same people. Iberian may just mean thatyou have a little Mexican mix Spanish and Indian., or may Portugeses. Depending on what service you use, you can get different populations.It would be good if you can get a taste from a relative that is not from the mystery family so you can separate the two.

 

[Squeek317]

I have tested with ancestry and downloaded my raw DNA data to aid my family research. Then uploaded it to GEDmatch.com.
I have found the site useful as a learning tool to educate myself on DNA in addition to "widening the net" to those seeking matches who have tested with other companies (ie 23andme, FTDNA X DNA , FTDNA Family Finder).

Through experience, I have found people who have gone through the trouble of the creating a GEDmatch acct as well as the DNA download/upload process are more likely to respond to emails regarding matches. It also gives you the option to attach a GEDCOM to your DNA.
Everyone has their reasons for doing testing, (myself= rare & difficult paternal ancestry to research, sis-in-law=adoption), IMO, folks using GEDmatch as a supplement seem more motivated.
I'm sure all this sounds like a plug, but as far as I know the site is run on donations. Just putting it out there as an FYI.

There is also a Chrome Extension for Ancestry DNA available for upload although I have no experience with it. It exists to make searching ancestry DNA matches easier & manipulating your matches in a database in different ways.
Hope this was helpful.

Mark

 

[Allie]

Day 2, things are going swimmingly. I added a whole new branch which I had overlooked an obvious paper link to before because a cousin match pointed it out to me, and about 20 of my "no idea who this is" matches turned into shared ancestor matches.

Also I may have a closer lead on my Woodsons. Another match popped up for a high confidence 4th cousin, and it turns out this line matches the profile I need based on what I already know from traditional census research: born in Virginia, lived in Kentucky long enough to have a child, died in Shelby county TN. And goes back to the Henrico/ Cumberland Virginia Woodsons I already had figured out in the 1600's. I haven't found the right individual yet but have a couple of theories.

 

[JPK Huson 1863]

Yes, Ancestry keeps adding them. It's surprising how distantly it'll pick up- I'll type in names from quite a few generations back to test the thing and get enough results to convince me. It's also being helpful trying to find one of our dead ends. All those names show up in one place- still can't find the connection, at least know where to look.

There was one line I'd had to make a good guess on pre-test- had around 8 common relatives show up with the same thing. It's great for verifying trees. We just did it to find loose ends- so far it's been worth it except for one maddening thing. There's this 22% Greek thing- we're not Greek, or did not think we were.

 

[18thVirginia]

I have tested with ancestry and downloaded my raw DNA data to aid my family research. Then uploaded it to GEDmatch.com.
I have found the site useful as a learning tool to educate myself on DNA in addition to "widening the net" to those seeking matches who have tested with other companies (ie 23andme, FTDNA X DNA , FTDNA Family Finder).

Through experience, I have found people who have gone through the trouble of the creating a GEDmatch acct as well as the DNA download/upload process are more likely to respond to emails regarding matches. It also gives you the option to attach a GEDCOM to your DNA.
Everyone has their reasons for doing testing, (myself= rare & difficult paternal ancestry to research, sis-in-law=adoption), IMO, folks using GEDmatch as a supplement seem more motivated.
I'm sure all this sounds like a plug, but as far as I know the site is run on donations. Just putting it out there as an FYI.

There is also a Chrome Extension for Ancestry DNA available for upload although I have no experience with it. It exists to make searching ancestry DNA matches easier & manipulating your matches in a database in different ways.
Hope this was helpful.

Mark

Thanks, Squeek. The Chrome Extension sounds really helpful. I have a Montgomery ancestor whom I have no idea about past when he arrived in Arkansas, supposedly from Tennessee. Since there are various Montgomery's and they tended to name their children the same names and two families of Montgomery's who started out in Virginia intermarried, it seems pretty difficult to find where your match in the family is. In fact, one branch of the family went through years of thinking they were part of another branch and DNA matching helped them figure out where their line started and who their ancestors were.

These were people with major genealogical experience, who'd done major research work, so it's interesting to see how DNA matching really helped their search. There's a Montgomery family DNA testing group, so there may be groups for other families as well.

I found the Montgomery family DNA testing group when I happened to see a message on some forum where some Montgomery descendant said to another, "Your ancestor wasn't in that group, your DNA doesn't match the Harold Montgomery group, see the testing group."

 

[lakertaker]

Without getting overly long-winded with details, Ancestry analysis indicated 62% United Kingdom. The UK includes England, Wales, Scotland, and Northern Ireland. That sure doesn't narrow it down too much.

 

[Allie]

On the other thread where we were discussing DNA results before, I mentioned that I wondered if my slave-owning ancestors had any children by the people they enslaved, and if any African ancestry matches would show up. Well, I now have an extremely high 4th cousin match of mixed European and Bantu ethnicity. Unfortunately no linked tree and it looks like this person probably just bought the test for the ethnicity results and has not been to the site in a while, so there's no way for me to know anything about where we link up.

 

[18thVirginia]

Here's a person who blogs about DNA and genealogy and has this blog post about how to use Ancestry DNA to help with your genealogical research. She talks about violating her own principles about making guesses and also about how DNA matches can help you determine if you're dealing with a "bad tree" or a "good tree."
http://dna-explained.com/2015/02/02/getting-the-most-out-of-ancestrydna/

 

[Allie]

Today I'm trying to get the hang of gedmatch. I understand that it's run on contributions, but couldn't they offer some undergrad student extra credit if he made something resembling an interface? I keep being shocked at how 1993 this site is. It would be way less of a load on their servers if you could search only for what you wanted, instead of having to search for everything and then narrow it down.

Also, I'm very annoyed with everyone who is related to me. So far I've found two dozen people who share the same huge match on chromosome 4 and not one of them has a gedcom to compare it to. Why would someone bother using this service without a gedcom?

 

[Allie]

Having used Gedmatch for a little while now, I have to say I still find it a very frustrating service. Having to copy and paste the same information over and over every time you want to look at something is exhausting, and not having a simple one-click interface to connect Gedcoms to DNA kits is annoying. Every time you leave the main page it logs you out and you have to log back in again, then it gives you an error at the login page, so you have to have the root page bookmarked and go to it. Seriously, if you're reading this and you run a college level computer class, contact these people and offer one of your kids extra credit if they will design a better site. Plus there are frequent outages.

Although the Gedmatch tools themselves are incredibly useful, I haven't gotten much value from them because my matches haven't uploaded trees, just DNA.

On the other hand, it's free, so I guess I've gotten my money's worth. If I can persuade some of my matches from Ancestry to upload their data there, it will be a lot more valuable.

Interesting development on the DNA match front: my closest match on Ancestry, a 3rd cousin, does not know who her biological father is and wants help with that. Should be interesting.

I also tried downloading the Google Chrome extension called Ancestry DNA Helper, a third party extension which is supposed to download a list of matches and ancestors of matches, convert these into database form, and allow you to use various utilities. This is broken right now. According to the support site for the extension, it has been broken since April, due to a Chrome update. According to them, trying to work around the bug can actually cause your computer to run its CPU constantly and overheat, so it's probably best to just avoid this one.

 

[BKey]

I am a Johnston researcher and "brick walled" in Bute County, North Carolina in 1761. I Y DNA'd a California cousin in 2006 with FTDNA because they have a Johnson, Johnston Johnstone sir name project. If I recall correctly, "we" perfectly matched four of five other "Johnsons" that had already tested. Their Johnson ancestor was born in North Carolinian in 1765.

So, after several more years of research, I think that I could prove in court that the father of my 1761 ancestor was a Francis Johnston (c. 1715 – 1790) and that the 1765 Johnson was a nephew of my 1761 ancestor and was born in the same area.

In order to find his origin, I then searched the records North to the James River and East to the Atlantic. I had no luck and wonder, "Where do I go from here?"

Since I tested, we have not had a single match, unless it was from a known family member or an unrelated name –Tucker, Lane, Bass, Baird, Berwick, etc. I think it is really odd that we have not had a Johnson / Johnston match in all of those years. So….I patiently wait for someone new to appear and give me a new lead.

I am a big fan of FTDNA because of their surname projects. The name of the game is for you to be able to find a match or they find you. Most everyone at FTDNA has their pedigrees at their project site and they are open to the public, which is a strike against Ancestry, IMO.

 

[Allie]

What DNA does and does not prove - not necessarily obvious to the typical user.

Unless you are descended from someone famous enough to have been dug up and had DNA extracted from his teeth or something, DNA matches don't "prove" anything about your relationship to an ancestor. They only prove that you do or do not match people supposed to have been descended from that ancestor.

But if they are wrong about their line of descent, suddenly all bets are off.

For example - I'm descended from someone who may or may not be the son of a famous man in the Colonial era. This famous man had a will which does not name my ancestor, but did name five daughters, and one son. My line has frequent matches with all the descendants of all five known daughters of this man. However, the official take away - stated on genealogical forums - is that our line has been proven not to be descended from him, because Y DNA from straight line male descendants of our ancestor doesn't match Y DNA from his named son.

Now, a little explanation about Y DNA. Y DNA is particularly useful in genetics because the Y chromosome, which causes maleness, is the only chromosome which is not part of a pair. Ordinarily when a parent donates half of their genes to a child, there is shuffling that goes on between the parents' paired chromosomes - genetic recombination. That's why full siblings who are not identical come out different every time. But since the Y chromosome has nothing to shuffle itself with, it remains exactly the same except for mutations, over many generations. The straight line male descendant of a man should have identical Y DNA to his ancestor.

So, in this case, my guy's Y DNA does not match the ancestor's Y DNA. If we were talking about autosomal DNA, the kind tested by Ancestry.com which comes from many parts of multiple chromosomes, this would mean nothing - there's no guarantee of any match at all to a distant ancestor, their genes can easily get completely lost in the shuffle. But Y DNA is different. Y DNA stays the same, so if there is no match, that's proof there is no relationship.

Only, there's a problem with the statements I made in the previous paragraph.

The problem is that we, today, are not comparing Y DNA to the Y DNA of this distant famous person. We're comparing it to the Y DNA of people believed, based on the paper trail, to be his descendants, based on the fact that his will indicates their ancestor was definitely regarded as his son.

But what if this famous man's wife had an affair? What if the reason our guy's DNA matches all the descendants of his daughters but not the Y DNA of this son is that this son was never his son at all?

In modern families, it's fairly easy to use DNA to sort out legitimacy claims. You can look at a genome and see if someone is a full sibling or a half sibling. That's not so easy to do ten generations in the past.

In this case, if you could look at straight male line descendants from a generation further back, you could figure out whether the son from the will shared his grandfather's Y DNA, and thus, presumably, his father's Y DNA. But the famous man's father had no brothers, and the paper trail becomes too fuzzy to be useful farther back. So there's no way of proving an actual biological relationship between the famous man and the son he named in his will. The named son's line shares matches with the man's daughters, just as our line does - but if he was illegitimate, or a son of his mother's known previous marriage, those matches could come through his mother.

There's no proof either way. Or, rather, the fact which is proved by DNA is, neither more nor less, that men whose paper trail indicates they are descended from possible son, our line, do not share Y DNA with men whose paper trail indicates the are descended from the son in the will, their line. Both groups equally lack the triangulation to prove through DNA that they are descended from the famous man himself. The descendants of the man in the will continue to have what they had before in terms of traditional genealogical proof - legal proof that their ancestor was regarded as the son of the famous man. Our guy's descendants continue to have the same traditional paper trail they had before, which is to say a lack of proof, but a likelihood that people sharing an uncommon surname in a small location might have something to do with each other, along with additional autosomal indications that they are related somehow.

What has been proven is that two modern sets of people do not share a common male ancestor. One set has a better traditional paper trail than the other. However, paper trails are not proof of genetic descent. Plenty of people whose birth certificates clearly state their fathers' names are in no way related to him. Modern DNA testing of large sample groups has found that roughly 2% of people who are sampled at random never having doubted their paternity are not the biological children of their fathers - in cases where there was enough of a doubt that someone sought testing, the rate is 30%.

So the correct phrasing here is, "Y DNA testing of my ancestor's descendants does not match the descendants of the son named in his will, so it's not probable he was this man's son." The frequently heard statement on forums, "YOU ARE NOT DESCENDED FROM HIM WE HAVE PROOF!" is not quite accurate, and the additional statement made by the descendants of the son in the will "We have DNA proof that we are descended from him and you are not," is simply false. They have traditional proof that they are descended from him, and DNA proof that their direct male ancestor was different from ours. They have no more DNA proof of straight male line descent from him than we do.

The example I've given may be a rare case, but the important thing to remember is this: your DNA matches or lack of same are to modern people, not to ancestors directly. The value of those matches is only as good as the traditional paper trail.

 

[Xonk]

Because my surname family paper trail doesn't go back past 1803 I did a Y-DNA test with FTDNA. I was able to connect with a large number of close cousins. I even found my family in Ireland, where we arrived in 1608 as soldiers with the first Ulster Plantation. You could say I have the 2 ends, but haven't made the connection in the middle,

But, as has been pointed out, not sll Y-DNA results come out as expected. One of my best friends family did the FTDNA test and instead of matching others in their surname project.... their family matches my family. The family is Irish Catholic and arrived here well after my mine. The early Y-DNA test suggest her family fits between me and my cousins in Ireland. More testing would be required, upgrading to at lest 64 markers, but her family seems to have lost all interest in pursuing the matter.

 

[Allie]

Ancestry DNA has a new and very useful tool. Under the DNA match is a button allowing you to view matches you share with each match. They should have done this a long time ago, but it's hugely useful. I've already sorted several cousins who I didn't know how they were related to me.

 

[BelleBlackburn]

I did the National Geographic thing, which was fairly vague, but I need to get some Maury Povitch style DNA done. My mother was adopted and I need to find out if her father was the same as someone else. Does anyone know who does that kind of definitive DNA?

 

[Northern Light]

What DNA does and does not prove - not necessarily obvious to the typical user.

Unless you are descended from someone famous enough to have been dug up and had DNA extracted from his teeth or something, DNA matches don't "prove" anything about your relationship to an ancestor. They only prove that you do or do not match people supposed to have been descended from that ancestor.

But if they are wrong about their line of descent, suddenly all bets are off.

For example - I'm descended from someone who may or may not be the son of a famous man in the Colonial era. This famous man had a will which does not name my ancestor, but did name five daughters, and one son. My line has frequent matches with all the descendants of all five known daughters of this man. However, the official take away - stated on genealogical forums - is that our line has been proven not to be descended from him, because Y DNA from straight line male descendants of our ancestor doesn't match Y DNA from his named son.

Now, a little explanation about Y DNA. Y DNA is particularly useful in genetics because the Y chromosome, which causes maleness, is the only chromosome which is not part of a pair. Ordinarily when a parent donates half of their genes to a child, there is shuffling that goes on between the parents' paired chromosomes - genetic recombination. That's why full siblings who are not identical come out different every time. But since the Y chromosome has nothing to shuffle itself with, it remains exactly the same except for mutations, over many generations. The straight line male descendant of a man should have identical Y DNA to his ancestor.

So, in this case, my guy's Y DNA does not match the ancestor's Y DNA. If we were talking about autosomal DNA, the kind tested by Ancestry.com which comes from many parts of multiple chromosomes, this would mean nothing - there's no guarantee of any match at all to a distant ancestor, their genes can easily get completely lost in the shuffle. But Y DNA is different. Y DNA stays the same, so if there is no match, that's proof there is no relationship.

Only, there's a problem with the statements I made in the previous paragraph.

The problem is that we, today, are not comparing Y DNA to the Y DNA of this distant famous person. We're comparing it to the Y DNA of people believed, based on the paper trail, to be his descendants, based on the fact that his will indicates their ancestor was definitely regarded as his son.

But what if this famous man's wife had an affair? What if the reason our guy's DNA matches all the descendants of his daughters but not the Y DNA of this son is that this son was never his son at all?

In modern families, it's fairly easy to use DNA to sort out legitimacy claims. You can look at a genome and see if someone is a full sibling or a half sibling. That's not so easy to do ten generations in the past.

In this case, if you could look at straight male line descendants from a generation further back, you could figure out whether the son from the will shared his grandfather's Y DNA, and thus, presumably, his father's Y DNA. But the famous man's father had no brothers, and the paper trail becomes too fuzzy to be useful farther back. So there's no way of proving an actual biological relationship between the famous man and the son he named in his will. The named son's line shares matches with the man's daughters, just as our line does - but if he was illegitimate, or a son of his mother's known previous marriage, those matches could come through his mother.

There's no proof either way. Or, rather, the fact which is proved by DNA is, neither more nor less, that men whose paper trail indicates they are descended from possible son, our line, do not share Y DNA with men whose paper trail indicates the are descended from the son in the will, their line. Both groups equally lack the triangulation to prove through DNA that they are descended from the famous man himself. The descendants of the man in the will continue to have what they had before in terms of traditional genealogical proof - legal proof that their ancestor was regarded as the son of the famous man. Our guy's descendants continue to have the same traditional paper trail they had before, which is to say a lack of proof, but a likelihood that people sharing an uncommon surname in a small location might have something to do with each other, along with additional autosomal indications that they are related somehow.

What has been proven is that two modern sets of people do not share a common male ancestor. One set has a better traditional paper trail than the other. However, paper trails are not proof of genetic descent. Plenty of people whose birth certificates clearly state their fathers' names are in no way related to him. Modern DNA testing of large sample groups has found that roughly 2% of people who are sampled at random never having doubted their paternity are not the biological children of their fathers - in cases where there was enough of a doubt that someone sought testing, the rate is 30%.

So the correct phrasing here is, "Y DNA testing of my ancestor's descendants does not match the descendants of the son named in his will, so it's not probable he was this man's son." The frequently heard statement on forums, "YOU ARE NOT DESCENDED FROM HIM WE HAVE PROOF!" is not quite accurate, and the additional statement made by the descendants of the son in the will "We have DNA proof that we are descended from him and you are not," is simply false. They have traditional proof that they are descended from him, and DNA proof that their direct male ancestor was different from ours. They have no more DNA proof of straight male line descent from him than we do.

The example I've given may be a rare case, but the important thing to remember is this: your DNA matches or lack of same are to modern people, not to ancestors directly. The value of those matches is only as good as the traditional paper trail.

Whew! So, what you are saying, I think, is that although in every legal sense, there is a direct connection to the "famous" person, in fact those mentioned in the will might not actually be direct descendents according to DNA, while your family although not in any legal sense connected, might be more directly connected, although perhaps on the wrong side of the blanket? Or have I completely missed the mark?

 

[Allie]

I did the National Geographic thing, which was fairly vague, but I need to get some Maury Povitch style DNA done. My mother was adopted and I need to find out if her father was the same as someone else. Does anyone know who does that kind of definitive DNA?

Either Ancestry or 23andme offers autosomal testing. I only have experience with Ancestry. For the sort of thing you're doing you would also want to download your raw data from Ancestry and look at it on something like Gedmatch which allows chromosome browsing. What service did you test with before?

 

[Allie]

Whew! So, what you are saying, I think, is that although in every legal sense, there is a direct connection to the "famous" person, in fact those mentioned in the will might not actually be direct descendents according to DNA, while your family although not in any legal sense connected, might be more directly connected, although perhaps on the wrong side of the blanket? Or have I completely missed the mark?

Yep, that's what I mean. Although in this case it appears possible both sides are legally connected, just that one is not mentioned in the will. Sometimes that can be for a perfectly legitimate reason such as one child having received their share during a parents' lifetime.