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 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.

Friday, September 18, 2015

How Phasing At 23andMe Improved My Results

My Mother's results were phased with mine a couple of weeks after her initial results came in.

My unphased results seemed to suggest my father had some Native American admixture. After phasing practically all of my Native America shifted to my maternal side. It does look like around 1% still remains with my paternal line. I've been trying to establish whether my father did have a Native American ancestor? It's possible we descend from the Half King Tanacharison? Probably impossible to prove or disprove a connection from 17th century Pennsylvania?

Pre phasing
I am part French Canadian and have had a number of matches from that line. It was interesting to see French/German percentages increase after phasing. I was 5.6% French/German before phasing. Now I'm 12.1%.  This is through my paternal side. I believe this percentage also includes my Burgenland, Austria ancestry. My Eastern European increased from 1.7% to 4.8% for my paternal line. The higher percentage is correct because my paternal Grandfather was Eastern European. I always thought that my paternal Grandfather was substantially Ashkenazi, but I see no evidence of that in our DNA results? Oddly I'm finding some Ashkenazi on my mother's side? I'm now showing a trace of Ashkenazi through my mother's line. I didn't have any before phasing.

This is what my phased paternal results look like.

My maternal phased line improved also after phasing. My mother was half Nicaraguan on her Mother's side, and mostly Scots-Irish and German on her father's side. The other DNA testing companies failed to find much Iberian admixture in our results. Our results at 23andMe, post phasing, show substantial Iberian admixture. I was 6% Iberian before phasing, and now I am 16% after phasing.
Here is my ethnicity prediction for my Maternal line.

Phasing against my mother improved the accuracy of our results. Instead of ethnicity being assigned to broad categories the results became more specific.
I'm also happy with the fact my matches have an M if they also match my mother.
Testing both parents. or a parent, if possible is very helpful. It's especially useful at 23andMe.


Sunday, September 13, 2015

Final 23andMe Results Come In Before Mother's Passing

My mother Edna Forgey-Kapple died on August 24, 2015. She had been  bedridden for a couple years. Her quality of life was poor, so her passing was a blessing. She tested with 23andMe about a month before she passed away. I wasn't sure if she would pass their test considering her declining health condition, but she did.

The initial result was the Neanderthal percentage. My Mom had a higher percentage of Neanderthal. She had 3.3 and I have 3.1.

Unphased X Mom Orange Native
American Blue European
The next result to show up was the ethnicity predictions. The initial predictions didn't seem accurate, when looking at the unphased chromosome view. This was very apparent when looking at the X chromosome.  As you can see in the image, above left, the unphased side representing my Grandfather, who was primarily Scots-Irish and German, was given some Native American. I have not found any Native American ancestors in his line. After our results were phased all of the Native American was moved to the side representing my maternal Great-Grandmother who
was Nicaraguan. This seems correct. My own results were also corrected. The Native American on my paternal X is now on my maternal side, which makes more sense. Our Iberian and Southern European percentage also increased substantially after phasing. This makes sense considering our Hispanic/Nicaraguan roots. According to AncestryDNA our Southern European is mainly Italian? This doesn't really make sense considering our Hispanic heritage. I know the ethnicity results aren't conclusive proof of anything, but I do feel like 23andMe's phased results are closest to correct.

Phasing has left a trace amount of possible Native American on both my father's side, and maternal grandfather's side. I do see possible Native American on my father's side, but none on my maternal grandfather's side.

Looking at the phasing that 23andMe did for ethnicity I have to say they were right on the mark when it came to separating out my mother's parents' ethnicities.

The ethnicity predictions also look good when comparing my Mom's two nieces and nephew. It's especially apparent when comparing on the X.

Two Nieces are green and blue. Nephew Purple
Shared segments with my Mom
Aunt's Ethnicity (my mother)
Niece 1 Ethnicity results
Niece 2 Ethnicity results
Nephew Ethnicity

Where Nieces (sisters) share with each other.
Light blue half identical. Blue fully identical.

It appears that my Mom, her Nieces, and Nephew inherited quite a bit of common DNA on the X. Niece 1 shares DNA along the entire X with my Mom. This is because she inherited nearly her entire X chromosome from her maternal grandfather. Since her mother was my Mom's sister, this could occur. Sisters inherit identical paternal X chromosomes, which is why Niece 1 shares DNA on the entire X.

Another conclusion we can draw is that the Nieces (sisters) share European DNA from their maternal Grandfather at the start of their maternal X. This is because this portion of my Mom's maternal X is Native American, and her paternal X at this position is European. It also supports the fact Niece 1 inherited almost her entire X from her grandfather.

The final results to come in were the match results. I found some close matches. I have been able to get the segment information for most of her close matches. Hoping the segment information may eventually prove helpful in breaking down some brick walls. I'm so glad I tested my Mom with 23andMe. She definitely lives on, along with all of our departed relatives, as can be seen in our common DNA.


Sunday, August 2, 2015

Using Public Matches From 23andMe To Map Segments

After receiving my 23andMe results I noticed some of my matches were listed as public. I assumed this meant I could see our matching segments without asking? I was wondering where their information was located? I didn't see it in Family Inheritance Advanced. After my Countries of Ancestry information became available, a couple weeks after I received my results, I noticed you could download a spreadsheet with your matches' segment information. This information seems to be from public matches? Some of those listed on the spreadsheet are listed by first and last name. Others are anonymous. Those listed with names, and those who are anonymous, all have some information about their family's countries of origin. So the info can be useful even if there are no names attached. You can get an idea of ethnic origins of the segments. I found that I could use the chromosome start and end points at Kitty's DNA Mapper, and Genome Mate, if I added 5 zeros; plus removed the decimal point. I just added the information to my Family Tree DNA CSV spreadsheet. Then uploaded the CSV spreadsheet to Kitty's DNA Mapper
This info is especially helpful if they list a place of origin other than the United States, or any other country with a large immigrant population. Finding Irish and Nicaraguan matches has been helpful. I was able to make a segment map using Nicaraguan matches' segment data, from Family Tree DNA and 23andMe. Now I'm able to see where I've inherited DNA from my maternal grandmother.
It's a very handy feature when you've found a common surname also. I found a Browning match who hadn't got in contact with me, on the list. Nice you can access this information even if someone won't agree to genome sharing.
Nicaraguan Segment Map

Saturday, July 18, 2015

Looking At Ethnicity And Inheritance On The X Chromosome


The chromosome view at 23andMe's Ancestry Composition is very interesting. It's especially interesting if you have an ethnically mixed background. It's difficult to separate the various European ethnic groups. If you're 100% European it isn't as useful. If you're wondering whether you have Jewish, Native American, African, Asian, etc. ancestry it is very helpful.

In my case my Grandmother was from Nicaragua, a place with a great deal of ethnic mixing. Most Nicaraguans are either more than half European, Native American, or African. I believe my Grandmother was a little more than half European. She was probably around a quarter Native American, based on my Mom's DNA results at AncestryDNA. I've been finding the X data at 23andMe to be informative regarding the ethnic mix on our maternal grandmother's side. I'm also beginning to understand the way the X recombines, or doesn't. Sounds like the X is more resistant to recombining than the other chromosomes. So it's possible that an unrecombined X could be passed on, as may be the case with a male cousin, and myself. I looked at the X sharing chromosome inheritance charts. I assumed that I could have inherited X DNA from most of the close lines. Looking at my X DNA I see that I didn't inherit DNA on the X from all close ancestors who could have contributed to my X. Instead I received a majority of the DNA on my X from my maternal Great-Grandmother and Great-Grandfather, based on the Native American, very little if any from anyone else. At least I'm unaware of my Maternal Grandfather having any Native American blood? I plan on testing my Mom to get a better idea of exactly who contributed the most to my X. If my chromosome is completely Native American and Southern European I may have gotten an entire X chromosome from my Nicaraguan Grandmother.

My X chromosomes. Orange is Native American, Purple African, Blue European.
However, looking at maternal cousins results the Southern European may be Northern European? In that case I probably did inherit some X DNA from my Maternal Grandfather. My brother and sister cousins also share segments on in my European region on the X. However one of my female cousin's shared regions is identified as Northern European while the same segment is Southern European according to her brother and a sister's results. The cousins with the Native American and Southern European results look more like my own, overall. You can see from the last X chart that brother and younger sister share half identical DNA across nearly the entire X chromosome (as shown by the green bar).

Orange is Native American, Purple African, Blue European. Brother and younger and sister

His results and the younger sisters results are starkly different than their oldest sister whose maternal and paternal X is practically completely Northern European. Unless some of her Northern European is actually Southern? It's 3 to 1 that our shared European DNA is Southern European. 

Differences when comparing two sisters
You can see by green bar they share DNA across the X

I noticed when I compared the female cousins to their brother the European and Native American lined up very well, when comparing side by side, looking at the share regions. When I compared the two sisters I was surprised the sisters matched all the way across the chromosome, but didn't share the same ethnicity predictions. I forgot that females received identical X chromosomes from their father. Fathers pass on their one X chromosome, unchanged, to their daughters. It appears they match across the chromosome; but, in reality its combined DNA from different lines. An expert on the subject of the X chromosome inheritance, Dr. Kathy Johnson, advised using either GEDmatch or 23andMe's Family Traits chromosome browser to see the half identical and fully identical regions (see ISOGG Wiki for examples of half identical and fully identical segments shared by relatives on the X). The fully identical regions mean  DNA is shared on both the Maternal and Paternal X. The sisters only share half identical regions with their bother because males have only one X and can't be fully identical with sisters. The first fully identical region is in a place where they don't match their brother (see blue chromosome below). The two sisters and brother share DNA where the sisters have their second fully identical segment. The half identical region for the two sisters, not shared by their brother comes from their father. We know this because this is a half identical region for the two sisters, which the brother doesn't share (remember the brother and younger sister do match there. So the older sister inherited more X DNA from her maternal grandfather who was Scots-Irish and German. The younger sister and brother inherited more X from their maternal grandmother who was Nicaraguan). Looking at half identical and fully identical regions is a great way to find out where siblings DNA is inherited from, since this is not shown in the advanced inheritance chromosome browser at 23andMe. 
This is from Family Traits Browser at 23andMe
Showing full identical dark Blue and half identical light blue

Here is a chart Dr. Kathy Johnson made for us.

Chart shows my 2 female and one male cousins ethnicity results
From Family Traits and the Ancestry Composition chromosome views 23andMe

 I have a few more distant cousins who match me on the X. One of them is also Nicaraguan and confirms that that region of the X is Nicaraguan. I have a couple matches on my father's side also. One is a 3rd cousin. Since fathers don't pass down their X to their sons we know this DNA doesn't come from our shared French Canadian line.  
The X is very useful when you have a question involving the X lines of inheritance. My X DNA has confirmed what the other tests seem suggest regarding Nicaraguan marriage patterns. It looks like males with European direct Y lines often married either Native American or African females. Seems like it was more acceptable, and probably because of the smaller female European population, males often married outside their own ethnic group. European females didn't marry out of their own ethnic population as frequently as males. The European males were probably financially better, off making them more attractive marriage partners.
I'm attempting to test my Mom with 23andMe. She is very advanced in age, and in failing health. Not sure if her test will pass? Keeping my fingers crossed that the test will pass. It would be great to phase my results with hers. I could then get a better picture regarding the ethnicity on my maternal X. Curious to see if I've inherited any DNA from my maternal Grandfather or is it all from my Maternal Grandmother? I also see a small bit of Native American on the chromosome which represents my father. Looking at the X lines on my father's side I am unaware of any family tradition regarding Native American in those lines? Another question I would like to find an answer for? I mailed my Mom's kit in yesterday. I'm following its course through the mail using USPS tracking.

X inheritance

Sunday, July 5, 2015

How Do AncestryDNA cM Numbers Compare?

With the mirrors to the other world at AncestryDNA finally cracking open a little some comparison between their results and those from Family Tree DNA, 23andMe, and GEDmatch can be made (if you've been watching the BBC series Strange and Norrell you'll understand that reference).

 I only counted segments 5 cMs and over when comparing.

Starting the comparison with a double 5th cousin who has tested with all 3 companies:
This cousin has results at the 3 major companies plus GEDmatch. In this case AncestryDNA seems to have removed the 9 cM segment on Chromosome 6. This same segment was removed from two other relatives of this match, also. From totals around 24 cMs with GEDmatch and the other two companies we go down to a total of 13.404 at AncestryDNA.

AncestryDNA total shared 13.399 cM Total

Family Tree DNA 23.41 cM Total

GEDmatch cM Total 25.2
23andMe cM Total 24.3

Moving on to this match who has tested with the 3 major companies:
In this case AncestryDNA cut the cM total from 25 to 26 cM's to 6.748. This would mean, if Ancestry is right, this may not be a true cousin match. I haven't established a relationship in this, case so far. So they may be right in this case?

AncestryDNA Total 6.748
Family Tree DNA Total 25.27

23andMe Total 26.2
Moving on to a 4th Cousin:
The particular match is a known, confirmed, 4th cousin. AncestryDNA predicts he is a 5th to 8th cousin. AncestryDNA totals are much smaller with only 6.748 total, while 23andMe says we share 32 cMs. I tend to believe the higher number.
AncestryDNA 4th cousin match total 6.748 cMs
23andMe 4th cousin match total 32 cM's
I've compared a couple more matches, which basically follow along these same lines. I haven't found any cM totals predicted to be closer than 8 cM's, when comparing with AncestryDNA. Some comparisons are 20 cMs different. I've made notes for all of the matches I've been able to identify at AncestryDNA. Looking at the totals for some of these matches I'm surprised at how low the cM totals are. I believe if we compared elsewhere the totals would be higher. I don't know about how the phasing and filtering is working for anyone else? It's not working well for me. I believe the best approach to matching is to provide testers with unphased, unfiltered matches who share a 7 cM segment or more, and with at least 700 SNPs. I don't feel the current matching system is helpful enough in eliminating bad matches, because there are many matches sharing cM totals under 7 (many sharing only 5 cMs total). AncestryDNA's system also tends to remove some good matches. Losing good matches, and having close cousins downgraded, in order to eliminate a few bad matches isn't worth the trade off.


Monday, June 29, 2015

Keys To 23andMe Success

You can see the additional info the 23++ extension adds

My relative matches came in last Thursday at 23andMe. That was a week after the initial Neanderthal, haplogroup, and ethnicity results. So far I'm enjoying my experience. I love the chromosome browser with its ability to check to see if my matches actually match each other. You can also compare with non matches if you invite them and they accept your invitation. At 23andMe you can't compare in the chromosome browser unless the other person accepts your invitation. So far I've had a good acceptance rate, considering I just started sending invitations a couple of days ago. I sent out at least a couple hundred invitations so far and around 25 people have accepted genome sharing.

I immediately found some matches I'm definitely related to. Three of my first cousins matched me of course. One of them shared a higher percentage of DNA than the average for a 1rst cousin. She shared 18%. She was predicted to be my aunt instead of cousin because of that. I found a 3rd cousin right away also. I already knew her, so it wasn't a surprise. 23andMe predicted her to be a 2nd cousin because she also shares more DNA with me than the average 3rd cousin. She shares 171 cM and 6 segments. I discovered 4 cousins just looking at posted trees.

Forgey Roller? On Chr 20
Moving on from the easy cousin finds I began trying to triangulate using the chromosome browser to compare those who accepted my invitations so far. I did find triangulation between my cousins, and I, and a woman who also has Tennessee ancestry. I discovered she also matched an Andrew Forgey and Anna Roller descendant. We have not found the common ancestral line yet. I don't think her tree is out that far?

Keys to success at 23andMe

  1. Other close relatives need to test with them. I've found it's so helpful that my 1rst cousins have tested with them. It's helping me to determine which side of the family matches match on. I don't think I would have much success without close relatives testing.
  2. The 23++ Chrome browser extension is helping me so much. It's a must have for me. 23andMe matches would be difficult for me to evaluate without the extension. The extension provides you with cM totals. I'm used to evaluating matches based on cM's rather than percentages, plus most everyone outside of 23andMe uses cM's as a measurement. This extension also highlights matches you've invited marking those who have accepted with a green box, and those who have not with a tan box (see top of page).
  3. Downloading and comparing matches. Since you can only compare 5 matches at a time at 23andMe it's good to download matches to excel compare there; or better yet compare with all of your matches from every company at Genome Mate, or create a segment map using Kitty's segment mapper.
The note system they have at 23andMe isn't as easy to use as at Family Tree DNA and AncestryDNA. I would like to see a better note keeping tool, which highlights where the notes are.

23andMe, like Family Tree DNA, doesn't have anything like Circles. The only way to find matches in common is to share genomes. The difference between the Circles and using a chromosome browser is you can actually prove a relationship with matches by comparing segments. Circles provide hints to possible relationships only, and these hints need to be verified.

I'm thrilled to be building up a catalog of segments using 23andMe. I appreciate the fact they provide that tool.

Sunday, June 21, 2015

Happy Father's Day! And 23andMe Admixture Results

Roberta Estes featured her Y DNA line in her Father's Day tribute post "Father’s Day – Tracking the Y DNA Line." These men represent my Y DNA lines.

My Father Robert John Kapple born in Chicago
My father Robert John Kapple and his father Rudolph Christian Kapple represent the YDNA haplo group  J-M172. This reflects an Eastern European origin. Rudolph Kapple was born in Southern Burgenland, Austria.
Paternal Great Grandfather Rudolph Christian Kapple born in Burgenland, Austria
My Maternal Grandfather Charles Lynn Forgey born in Jackson County, Indiana

My Grandfather and his male ancestors were Y I-126 Haplo. This line has Scottish roots.
Great-Grandfather William Wray Forgey born in Jackson County, Indiana
Great-Great Grandfather Hugh Forgey probably born in Scott County, Virginia
23andMe Ethnicity results-
I received my preliminary 23andMe results this week. I found out I'm 3.1% Neanderthal. I've been reviewing my ethnicity results. I found that 23andMe slipped up on the Eastern European estimate. They only estimated me to be 1.70% Eastern European, which I believe is too low. My Aunt on my father's side is 29% Eastern European according to her myOrigins results. My Grandfather Kapple was born in Eastern Europe. Both AncestryDNA and myOrigins estimated I'm around 7% Eastern European. I believe even that estimate is low. Perhaps the 11% Broadly European may represent Eastern Europe? I was happy to see some French and German admix included in 23andMe's estimates. I believe their estimate should probably be higher also. I believe they also underestimated the Middle Eastern admix. My original Family Finder ethnicity results stated I was 10% Middle Eastern. They cut that estimate in half now. When my Broadly Southern European is added to my Iberian result it comes out to nearly 14%, similar to AncestryDNA. This would represent my Grandmother Graciela Del Castillo's family. It's fairly close to the 12% AncestryDNA predicted. The 3% from myOrigins is definitely too low. MyOrigins agrees with 23andMe giving me an overall 91% European. MyOrigins and 23andMe also agree that my British Isles admix is around 29%, which I believe is closer to correct than AncestryDNA.
AncestryDNA's ethnicity results are very good. They do, however, give a wide range of possibilities for each result. MyOrigins gets some regions correct, but misses large chunks of my admix. 23andMe throws a good percentage of my DNA into "broad European" categories. They won't breakdown the regions to the degree AncestryDNA does.
I'm currently waiting for my match results at 23andMe. I'm hoping some 2nd Cousins show up. I have no 2nd Cousins at Family Tree DNA or AncestryDNA. That cousin level could prove useful in order to separate the segments of DNA I share with more distant cousins.