Tuesday, October 20, 2015

The DNA Detectives Series

I noticed Kelly Wheaton's post (Facebook ISOGG) about "The DNA Detectives" a New Zealand based, genetic genealogy themed, TV show. Great idea for a show! Hopefully it will inspire more people to test. I have a few New Zealand and Australian matches; mostly based on my Irish ancestry.
You can watch the show here (see bottom of page) or at this link https://www.youtube.com/watch?v=IvFAyWoYpaY

The shows host is a little quirky. His animated gestures are funny, like pretending to blow a train whistle at one point, and making a train sound effect. The set in the studio is retro themed, and the host slams a button to display the DNA results. Two celebrity guests were featured, Jack Tame and Ray McVinnie, in this 44 minute show. After an initial ethnicity test reveal, in the studio, the guests travel around the world in search of DNA cousins.

South Asian segments
I enjoyed the show. I thought the first guest's, Jack Tame's, initial trip to New York was a let down, and I wasn't sure if the rest of the show would be similar (don't know why a painting couldn't be briefly unpacked and shown to him?). After the initial New York fail the rest of the show became increasingly interesting. I found the Elvis connection to the second guest, Ray McVinnie,  a little bit of a stretch. This guest was related to a family in Oklahoma, who in turn were 5th cousins to Elvis. Not a very close tie to Elvis. The fact the family was related to the judge who overturned the Rosa Parks ruling, ending segregation in the south, was more interesting to me. One of the women explaining the family history to McVinnie misspoke, saying he is related to the Federal Judge Frank Thomas, actually his name was Frank Johnson (it's easy to see how someone can misspeak leading to later confusion when seeking to find more info).

Kelly Wheaton's segments were interesting because they included more historical information, and actually showed the basis for one of the ethnicity percentages. Kelly had researched Tame's line. She found a surgeon, from Scotland, who immigrated to India in the 17th century. He married a native of India, which is where the 1.4% South Asian comes from. When Kelly showed him the large South Asian share from the 23andMe's ethnicity chromosome browser it really brought that fact home to him. He was blown away when he saw that. He had heard about some possible South Asian, but actually seeing the segment made it real. Kelly Wheaton is related to Tame, according to DNA test results, but she was not able to find the common ancestors. These distant cousins did share a resemblance as they noted.

The segment in Jamaica was very interesting also. This took McVinnie to Jamaica in search of the African ancestor who contributed the 2.2% African admixture. He met up with some distant cousins who showed him their family tree. Even though the guest wasn't aware of any connection with Jamaica he shared DNA and surnames in common, which established a definite relationship. The African ancestor was identified, and it sounds like he lived in the 18th century. He didn't expect his trace of Jewish ancestry to also stem from Jamaica, but it did. Portuguese Jews fled to Jamaica during the inquisition and other times of persecution.

The final segment featured someone only identified as James. This man designs software to help adoptees find relatives. He is an expert at hacking into computer systems. He doesn't hack illegally, he does it as a career to help companies find holes in their security. Great to have someone as skilled as he is helping adoptees!

Some educational information coming out of this show:
  1. Guest Tame's grandmother already knew some of the information which he discovered on his New York trip. This is something that always comes up on these kinds of shows. Relatives will later tell guests they already knew something. Many relatives never share stories unless asked.
  2. Small 1 and 2 percent DNA shares can go back to the 18th Century. The host said these small shares can go back 200 to 300 years, and are not ancient. Likely correct  The featured guests had small ethnicity segments of from 0.40% to 2.2% and these shares went back to the 18th Century.
  3. Guest two, McVinnie, would seem to have more Italian ethnicity than showed up in his ethnicity results. Southern European is not as well defined as Northern European when it come to these tests. Probably because most testers are of Northern European heritage.
  4. Always assume information passed down through generations could be wrong until you can find actual documentary evidence to support it. As a misspeak by one guest demonstrates.  
This is a fun show and I hope we see something like this in the US.

Monday, October 12, 2015

Has Our Mozabite Gone To DNALand?

I've been trying out the Beta version of DNA Land and, a new site devoted to scientific genetic research. It also provides some of the same features as the DNA testing companies; such as ethnicity results and matching for those upload their raw data file. Sounds like a very worthwhile project to contribute to. I encourage everyone (especially AncestryDNA customers) to contribute their raw data. Register and upload here https://dna.land/ Full instructions on how to get your raw data, and how to upload it are provided after you register. It's free to use.

When I got my initial results back I thought the Ashkenazi on my father's side had finally been recognized. My Ashkenazi came out to be 17%. After uploading my mother's raw data I discovered that she tested as 18% Ashkenazi/Levantine. A cousin thought that this result might reflect our shared German line. Looking at our family tree our German line represents a small fraction of our ancestry. I don't see it adding up to very much percentage wise. Plus most cousins in this line aren't testing Ashkenazi. At 23andMe my Mom and I have less than 1% Ashkenazi. None of the other companies found any Ashkenazi for either of us (except AncestryDNA did find a trace amount for me only).

For background I should say my maternal Grandmother Graciela Del Castillo was Nicaraguan, and my maternal Grandfather Charles Forgey was primarily Scots-Irish with some German. 23andMe found that trace of Ashkenazi, which it placed on my Grandmother Graciela's side after, in my Mother's results, after phasing. I believe all of the Middle East and Ashkenazi results are from my maternal Grandmother, since these results don't show up in Forgey cousin admixture.

Original Family Tree DNA result Mom
I remember back when my Mom and I first tested with Family Tree DNA we had high percentages of Middle Eastern. In my Mom's case it was said to be Mozabite. It's interesting that her Middle Eastern percentage (from 2013 FTDNA) matched her Ashkenazi/Levantine at DNA Land percentage exactly, at 18%. Family Tree DNA has changed that result to 8%. With 6% Asia minor (quite ancient ancestry probably?) and 2% North African. The 6% Asia Minor probably accounts for some of the 18% Ashkenazi/Levantine. I believe the change in our results at Family Tree DNA is due to the fact they separated out the Mediterranean from the Middle Eastern. Sea travel in the Mediterranean resulted in a great deal of mixing of peoples along the coastline. Also my Spanish ancestors likely lived in the Arabian occupied area of Spain. Spain also had a large Jewish population, which may account for a portion of the results? The Ashkenazi/Levantine results for my Mom and I would not trace to Eastern and Central Europe as shown on the map provided by DNA Land. They are probably tracing back to the Middle East and Asia Minor; farther back in time. I think that Family Tree DNA's 8 percent Middle Eastern is close to correct.

Since DNA Land is in early Beta the percentages might change? I have a feeling Mediterranean roots are difficult to untangle, and are mixed with North African and Middle Eastern. Our Mediterranean percentages are all over the place.

Looking at my Mom's Native American percentage prediction at DNA Land it looks correct. Native American is much easier to distinguish than some other admixtures.

I'm hoping DNA Land will provide information regarding the populations they are using to come up with the percentages? That would be helpful.

PS Spent the day at the beach yesterday. Got sunburned. I'm a little zonked out today. Hope this makes sense?

Friday, October 9, 2015

Using Segment Data In Evidence Citation

Yesterday the need for segment information was discussed at the Facebook ISOGG group, and also brought up by a Kapple DNA match. Many people say triangulation isn't working for them so the fact Ancestry doesn't provide segment information doesn't affect them. Since we all can agree that we can't accept undocumented facts and add them to our trees without substantiation, we should also agree we can't accept a DNA match without the segment information to provide evidence that we really do match, and where on the chromosome we share DNA. This is a form of citation and documentation. I appreciate Ancestry for the hints their DNA product provides, and the resulting collaboration. Occasionally I have been able to persuade matches to upload to GEDmatch from Ancestry. Ancestry isn't required to provide this information. If, and when, they do it will have to be done through permission from both parties agreeing to share segment info. This isn't an ideal situation, but more people would agree to share if they could just click a button and do it. I've had problems downloading the raw data file at times. This feature is sometimes down. It would be so much easier if you could just share the segment info at Ancestry. Ancestry isn't a strictly DNA company. As many have said most users wouldn't know what to do with the segment information if they had. This may be true now, but I believe many will have their curiosity piqued and eventually would use this information.

Thinking further about the importance of segment information as supporting evidence I feel I need to add the exact segment information to the ancestors the segments likely correspond with. I will add the AncestryDNA info to my genealogy software too, but I can only say that Ancestry predicts such and such relationship, but no supporting chromosome information is available. Saying "Ancestry says" isn't the best supporting evidence, unfortunately.

I've decided to add the segment information to notes. I noticed you can add Y and MtDNA information at Rootsmagic. I don't see a way to add segment info? I don't have the most recent version of the software. Maybe more recent versions have more options for adding DNA info? Somehow all of this DNA information needs to be integrated with the rest of my evidence.

We need substantive information from DNA testing companies in order to get the most out of our tests. We need substance so we can more confidently collaborate with our matches. We need segment information if we want to use DNA testing results as serious supportive evidence.