(Not-So-) Wordless Wednesday: GEDmatch Eye Color Prediction - Mom

Did you know that if you upload DNA test results to GEDmatch, you can use an eye color prediction tool on their site?

According to the site, the eye color predictor works best with DNA results from 23andMe.  In addition, the utility "used samples from the GEDmatch database which consists mostly of European descendants. As such, it will be less accurate for other ancestries."

So, knowing this in advance, I ran the predictor on my mom's DNA test results.  I did her autosomal test with Ancestry.com, then uploaded the raw data to FamilyTree DNA, and then ran a separate mitochondrial test on a new sample there. I'm not sure if this predictor uses autosomal or mitochondrial DNA or both.

Mom is of mostly European extraction.  Here is the result:

The results page also provides the following information:

"Read rules from top to bottom. In some cases, a rule cancels out results from rules above it.

CT at: rs17762363 - Increased melanin production. Adds yellow, amber, or brown. Some darkening. Contributes to brown.
CC at: rs3794604 - Blocks some melanin. Often gives light colored eyes.
GG at: rs7174027 - Blocks some melanin. Often gives light colored eyes.
CC at: rs4778241 - Low Melanin. Basis for Gray, Blue, Green, or Yellow Eyes if no other pigmentation is present.
CT at: rs3947367 - Contrasting sphincter around pupil.
TT at: rs1129038 - Penetrance modifier. Blue.
GT at: rs1470608 - Medium melanin on Anterior Epithelium. Gives dark eyes.
AG at: rs11634406 - Flecks (Nevi).

8 rules were used to make this prediction. There are 61 active rules in our evaluation model, utilizing 41 SNPs.

Please note that your FTDNA Illumina results do not contain many of the SNPs used by this utility, and therefore a considerable amount of accuracy is lost."

And here is my mom's eye in a photograph from February 2017:

Another picture of her eye, this time from Thanksgiving Day, 2016.  I don't have any color photos that clearly show her eyes from her younger days.

The website asks you to rate the accuracy of the prediction.  Here are the choices:

• It's exactly right.
• Color is correct. It missed one or two tiny details, but it's very close.
• Came close on the color(s), and got a lot of the details.
• Got some of the colors in my eye, but missed a lot of details.
• The color is not exactly close, but it's not a complete miss either.
• The prediction missed the color completely, but it picked up a few other details.
• Completely wrong. The color's not even close.

So what do you think?  Let me know in the comments!  I'm thinking this one is pretty close.


© Amanda Pape - 2017 - click here to e-mail me.

(Not-So-) Wordless Wednesday: GEDmatch Eye Color Prediction - Dad

Did you know that if you upload DNA test results to GEDmatch, you can use an eye color prediction tool on their site?

According to the site, the eye color predictor works best with DNA results from 23andMe.  In addition, the utility "used samples from the GEDmatch database which consists mostly of European descendants. As such, it will be less accurate for other ancestries."

So, knowing this in advance, I ran the predictor on my dad's DNA test results.  I did his autosomal test with Ancestry.com, then uploaded the raw data to FamilyTree DNA, and then ran a separate yDNA test on a new sample there. I'm not sure if this predictor uses autosomal or yDNA or both.

Dad is of mostly European extraction.  Here is the result:

The results page also provides the following information:

"Read rules from top to bottom. In some cases, a rule cancels out results from rules above it.

CC at: rs3794604 - Blocks some melanin. Often gives light colored eyes.
GG at: rs7174027 - Blocks some melanin. Often gives light colored eyes.
CC at: rs4778241 - Low Melanin. Basis for Gray, Blue, Green, or Yellow Eyes if no other pigmentation is present.
CT at: rs3947367 - Contrasting sphincter around pupil.
AA at: rs1105879 - Weak Amber Gradient
TT at: rs1129038 - Penetrance modifier. Blue.
AG at: rs10467971 - Penetrance Modifier - Blue
GG at: rs12906280 - Gray ring around outer edge.

8 rules were used to make this prediction. There are 61 active rules in our evaluation model, utilizing 41 SNPs.

Please note that your FTDNA Illumina results do not contain many of the SNPs used by this utility, and therefore a considerable amount of accuracy is lost."

And here is my dad's eye in a photograph from April 1957:

And just to show that it's not an issue with the color being off from a 60-year-old photo, here's a picture of my dad's eye from December 2008 (I didn't want to use just this one photo because of the glasses and the red eye reflection):

The website asks you to rate the accuracy of the prediction.  Here are the choices:

• It's exactly right.
• Color is correct. It missed one or two tiny details, but it's very close.
• Came close on the color(s), and got a lot of the details.
• Got some of the colors in my eye, but missed a lot of details.
• The color is not exactly close, but it's not a complete miss either.
• The prediction missed the color completely, but it picked up a few other details.
• Completely wrong. The color's not even close.

So what do you think?  Let me know in the comments!  I'm thinking this one is completely wrong.

© Amanda Pape - 2017 - click here to e-mail me.

(Not-So-) Wordless Wednesday: GEDmatch Eye Color Prediction

Did you know that if you upload DNA test results to GEDmatch, you can use an eye color prediction tool on their site?

According to the site, the eye color predictor works best with DNA results from 23andMe.  In addition, the utility "used samples from the GEDmatch database which consists mostly of European descendants. As such, it will be less accurate for other ancestries."

So, knowing this in advance, I ran the predictor on my husband's AncestryDNA test results.  He is of mostly European extraction.  Here is the result:

The results page also provides the following information:

"Read rules from top to bottom. In some cases, a rule cancels out results from rules above it.

GG at: rs7174027 - Blocks some melanin. Often gives light colored eyes.

CC at: rs4778241 - Low Melanin. Basis for Gray, Blue, Green, or Yellow Eyes if no other pigmentation is present.

CC at: rs9782955 - Blocks some melanin. Often gives light colored eyes.

TT at: rs1129038 - Penetrance modifier. Blue.

GG at: rs12906280 - Gray ring around outer edge

CC at: rs7403602 - High density on Anterior Stroma. Blocks melanin. Blocks blue. Gives lighter colors.

TT at: rs1667394 - Starburst (Collarette)

CC at: rs12203592 - No pigmented Collarette.

8 rules were used to make this prediction. There are 61 active rules in our evaluation model, utilizing 41 SNPs."

And here is my husband's eye in a photograph that I took in May 2016:

The website asks you to rate the accuracy of the prediction.  Here are the choices:

• It's exactly right.
• Color is correct. It missed one or two tiny details, but it's very close.
• Came close on the color(s), and got a lot of the details.
• Got some of the colors in my eye, but missed a lot of details.
• The color is not exactly close, but it's not a complete miss either.
• The prediction missed the color completely, but it picked up a few other details.
• Completely wrong. The color's not even close.

So what do you think?  Let me know in the comments!

© Amanda Pape - 2017 - click here to e-mail me.

Surname Saturday: Peach - DNA Discoveries

A while back, I had my husband's autosomal DNA tested with Ancestry.com. Then I downloaded the raw data file and uploaded it to FamilyTreeDNA, paying the $39 to unlock all the matches.

I've been contacted by a few people who are matches, and have been able to add some people to the family tree (and confirm other relationships) as a result.

One case is in my husband's Peach line.  His great-grandmother was Angelina Elizabeth "Lizzie" Peach Moore, who was born March 31,1859, in Salem, Lee County, Alabama, and died September 4, 1924, in Lewisville, Denton County, Texas.  She and her sister Emma Ann Peach Moore (1861-1915) married half-brothers who moved to Denton County in late 1882 or 1883.

Lizzie and Emma were the two oldest children of Jonathan A. Peach (born July 30, 1832, in Clarke County, Georgia; died March 21, 1908, in Concord, Blount County, Alabama) and his wife Angeline "Ann" Emily Johnson Peach (born MAY 1836 in Alabama, died January 18, 1921, most likely in Lee County, Alabama).

Jonathan, in turn, according to the 1850 Census, was the son of James Peach (born about 1786, died after 1870) and his wife Louisa (born about 1790, died after 1860 but probably before 1870).  On the 1850 Census, Jonathan is the only child listed, so I was not aware he had any siblings.

About a week ago I was contacted by a gentleman who manages the DNA accounts for his wife and three sons.  He said my husband was a match to the wife and two of the sons.  I used the Family Finder Chromosome Browser in FTDNA to take a look.  Here are the matches - the wife is orange, the two sons are blue and green.  Let's call them Orange Wife, Blue Son, and Green Son.  The matches are on the second and fourth chromosomes.  My husband also has a small match with Blue Son on the 15th chromosome (not shown).

As it turns out, Orange Wife also has James Peach as a 3rd great-grandfather - meaning she and my husband are fourth cousins.  James Peach apparently had 11 children between 1808 and 1832, with my husband's 2nd great-grandfather as the youngest.  Orange Wife, Blue Son, and Green Son are descendants of the fourth child in the family, George Peach (1813-1899).

The gentleman (Orange Wife's husband and the father of Blue Son and Green Son) put me in touch with someone else who is apparently also kin through this Peach line.  This cousin (and her sister) match Orange Wife and Green Son on different chromosomes than my husband does, but have no match to Blue Son.  However, they have matches to Orange Wife's third son, who my husband does not match.

This cousin encouraged me to upload the raw data to GEDmatch.  I'd been asked to do this by yet another of my husband's matches (this time in AncestryDNA) and had not gotten around to it.  It turned out to be quite easy.

Here's some more data on my husband's matches:

My husband and Orange Wife DO appear to be a match of genealogical significance, using the one-to-one comparison tool in GEDmatch, and increasing the minimum SNP threshold from 500 to 700:



My husband and Blue Son are also a match on chromosome 2:



They also match (albeit much smaller) on chromosome 4:



My husband also matches Green Son on chromosome 2:



There is an overlap for my husband, Orange Wife, Blue Son, and Green Son on Chromosome 2 from 78,790,630 to 96,111,193.

I'm still learning how DNA works, but I feel pretty confident that my husband, Orange Wife, Blue Son, and Green Son are all related, and thus I have some more people to add to his Peach tree.


© Amanda Pape - 2016 - click here to e-mail me.

Surname Saturday: Ancestry.com Ethnicity Estimates for Mom & Dad

The offspring just asked some questions about where the ancestors came from, so here's a quick reply.  A while back I had my parents' autosomal DNA tested through Ancestry.com (and also uploaded the raw data files to Family Tree DNA).  Here are the Ethnicity Estimate charts for Dad and Mom:

 Dad's above and Mom's below.

Keep in mind that these charts reflect migration patterns from thousands of years ago.  That being said, the charts do fit what I know about the ancestry of my parents.  Dad's immigrant ancestors (Pape, Kramer/Cramer, Massmann, Dienes) came almost entirely from Germany (Europe West) - and of course, many of the Germanic peoples were originally from Scandinavia.  I have to say, I am intrigued by the Iberian Peninsula and Italy/Greece strains in his background.

Mom's paternal ancestors (Guokas, Banavich/Bonewitz/Boenewitch) immigrated from Lithuania, which is reflected in her Europe East and Finland/Northwest Russia ethnicity.  Her maternal ancestors included German immigrants (Wolfe, DeHaven, Horein - the Europe West and Scandinavia) as well as Shelton, Bridges, and Spikes, who likely originated in Great Britain and Ireland.

You can learn more about AncestryDNA here.

© Amanda Pape - 2014 - click here to e-mail me.