How to find out about DNA surprises?

Question

How to find out about DNA surprises?

1.0k views

I need help with a DNA question. My Aunt is not showing up as a "full" Aunt (father's sister) in my DNA matches over on Ancestry. I have uploaded both mine, hers, my sister, and her daughter's raw DNA to GEDmath, run a bunch of reports and still can't figure out how/why she is coming up as a 1C1R or 1/2 Aunt. We share 630cM.

As of for our family tree, I have all the info but have no clue where to even look. What is freaking me out is that she matches, very low at 104cM, with my Aunt on my mothers side too.

Anybody have advice as to what my next steps are? I am not telling her anything, she is 86 and in bad health.

Thanks,

Sheila

asked Jul 9, 2022 in Genealogy Help by Sheila Tidwell G2G6 Mach 6 (63.9k points)

4 Answers

Best answer

Well barring some weird malfunction at ancestry it's obviously impossible for your father and your aunt to be full siblings. In fact they could even be first cousins rather than half-siblings. You could use a second testing site but it's very unlikely to give a significantly different result.

The chart about two-thirds of the way down this page https://isogg.org/wiki/Autosomal_DNA_statistics is extremely useful as it gives an average value of shared DNA for each relationship plus a range. From the chart, half-aunt has a range 492-1315 (average 871) and 1CIR has a range 102-980 (average 433) so half-aunt is the likelier scenario.

You will have one X chromosome that's a perfect copy of your father's, which is a blend of his mother's two copies. Half of your aunt's X-DNA should also be a blend of these two copies. X-DNA is tricky but if you share at least 25%+ X-DNA with your aunt then I think it's very likely that her mother was indeed your grandmother. However, a lower value doesn't prove the opposite. In the most extreme case possible, your father could have inherited one copy and your aunt the other, so they would have no matching X-DNA.

answered Jul 10, 2022 by Matthew Fletcher G2G6 Pilot (133k points)
edited Jul 10, 2022 by Matthew Fletcher

Human's recieve 23 pairs of chromosomes or 46 total chromosomes (because they are paired). Half of the chomosomes come from each parent. But the way DNA is transfered, one does not recieve exactly half of each parents DNA. This is important to remember.

When the parents donate their 22 non-sex chromosomes (called Autosomes) each pair recombine DNA in a random manner, thus while we get half of each parent's chromosomes, we only get approximately half of our DNA from each parent because of recombination. That slight varriation makes a difference with each generation. And it is reflected in the autosomal DNA tests as it gets tweaked each generation.

Now the sex chromosomes. One from each parent.

The 23rd chromosome determines sex. X from Mother and Y from Father and you get a male. X from Mother and X from Father and you get a female. So XY male, XX female.

From his father, your brother being male, will get one copy of his father's Y chomosome (a chromatid) and approximately 50% of random autosomal DNA (one half of each autosome).

In other words. He gets a sex chromosome (Y - because he is male) and a random mix of his father's 22 autosomal DNA. These pair with his mother's X (sex) chromosome and her 22 autosomal DNA donation.

So...

Everyone gets autosomal DNA from both parents. A random mix of DNA from 22 chromosomes (non-sex chromosomes).

Females (XX) get a combination of X chromosomal DNA. One from her father (exact copy) and one from her mother (A mix of her mother's two X chromatids).

Males (XY), your brother, gets one Y chromatid (a copy of his father's) and one X chromatid from his mother. (This is also an exact copy of his mother's).

Not mentioned is Mitochondrial DNA which only comes from mothers. It is not chromosomal DNA but rather from the mitochondria from the mother. So both males and females inherit Mitochondrial DNA but only females pass it along.

Now, with regards to testing.

Y-DNA is testing the Y-Chromosome with is only found in Males. Females need to find an appropriate male relative if this test is needed.

Mitochrondial DNA (non-chromosomal) testing is for both males and females who wish to persue a maternal line as this is only passed along through females.

X-DNA testing is for both males and females but you must understand how the X chromosome does or does not recombine depending on which sex inherits this chromosome. Recombined in females and exact copies in males. Thus a specific inheritance pattern needs to be followed.

Autosomal DNA is for everyone. (test everyone!)

A combination of these tests can be very useful in finding and confirming relatives. Which combination depends upon your goals.

However. All DNA testing is very dependent on the well crafted trees of all participants. Are they well documented and deep?

DNA testing relies upon sound trees to produce useful results.

I hope this helps.

commented Jul 10, 2022 by Michael Stills G2G6 Pilot (530k points)
edited Jul 10, 2022 by Michael Stills

Just sneaking in for a tiny bit of clarification, Michael...

"When the parents donate their 22 non-sex chromosomes (called Autosomes) each pair recombine DNA in a random manner, thus while we get half of each parent's chromosomes, we only get approximately half of our DNA from each parent because of recombination."

If we remove the sex chromosomes from the equation and look only at the autosomes (and barring any genetic structural abnormality like aneuploidy) we get exactly 50% of our DNA from each parent. It can't happen any other way because the gamete each generates contains a full, haploid chromosome.

I think where some confusion might come in is that recombination, technically "crossing over," doesn't happen when the parents' haploid chromosomes pair up after fertilization. Crossing over happens before the gamete is ever produced. In fact, in women, crossing over happens in all the oocytes the mother will ever have while she, herself is still in the womb.

It takes place during Prophase I of meiosis, and is what makes daughter cells genetically unique from one another. From there, the process enters Metaphase I, then Anaphase I, and Telophase I. Crossing over can't happen again after Prophase I is complete. Meiosis repeats those stages as Prophase II, Metaphase II, and so forth before final creation of the gamete. In the oocyte, meiosis is paused after Prophase I until puberty, and again at Metaphase II until the ovum is fertilized.

So we do get exactly half our autosomal DNA from each parent, but what we receive won't be equal 25% parts from our grandparents. That's when the variability starts.

That's also why, other than identical (monozygotic) twins, no two siblings will have the same autosomal DNA. The way 23andMe reports sibling sharing, they state the possible range as being from 38% to 61%. For a little fun with math, if we say that full siblings will share on average 50% of their half-identical region (HIR) DNA, we can estimate with a simple equation, in aggregate and roughly, how much of the parents' genomes will be passed down based on the total number of children they have: 1-(0.5ⁿ) where n is the total number of children. So two children would account for about 75% of the parents' autosomal DNA; four children, 93.75%; and we'd need seven children to get over the 99% mark.

commented Jul 10, 2022 by Edison Williams G2G6 Pilot (447k points)

Answer 1 · 2022-07-10T07:53:22+0000

I think your next step is to look for people she matches but you wouldn't expect her to match, and for people she should match but doesn't.

You've already made a good start on that, noticing that she matches your maternal aunt. As Michael Stills says below, you'll need well-developed trees to interpret that. How confident are you that the match can't be explained by a common ancestor of your two aunts? How far back, and how well documented, are their trees? But look also for other matches that don't make sense in light of her relationships on paper.

Also, are there people (preferably a cluster of people) she doesn't match whom you'd expect her to match? (First and second cousins should match, but a small percentage of third cousins don't. See ISOGG's 'Cousin Statistics' for more: https://isogg.org/wiki/Cousin_statistics )

You might find DNA Painter's "What are the odds?" tool useful. It lets you put in the amount of DNA people share, and test various hypotheses about the biological relationships. It's available here: https://dnapainter.com/tools/probability

answered Jul 10, 2022 by Harry Ide G2G6 Mach 9 (93.5k points)

I'll add to Harry's excellent answer that DNA Painter itself is often helpful in research and documentation. Jonny Perl continuously adds features and, for me, it's become a must-have. I've had a paid subscription since they were first offered which allows me, among other things, to maintain up to 50 profiles (since we use the term "profile" differently here, I tend to refer to them as "canvases" when talking WikiTree) and to have unlimited bulk import functionality to rapidly enter testing data. You can also opt to share any particular canvas with others via a unique, unpublished link. Very handy when you have an active collaborator working with you on a family line.

A good place to start if you want to look into DNA Painter as a possible arrow in your quiver--so to speak--is https://dnapainter.com/help.

There's nothing we can do about AncestryDNA not providing us with any chromosomal detail, so I still maintain a good ol' fashioned spreadsheet that stores all the data, and I use the same (or close to) colors in Ancestry to keep DNA Painter and AncestryDNA visually synched. For Ancestry users, here's a good article about color schemes there by Leah Larkin.

I use Leah's basic color scheme, except that I don't use a color as the profile picture in the Ancestry tree. (Hint to Ancestry.com: give us another option--a line under the profile picture, another item on the same line as the DNA connection buttons, or maybe best, a colored frame around the profile picture that will still appear in the tree view--so that we can use the identifying colors from AncestryDNA in our Ancestry.com trees.) The limit at AncestryDNA is 24 different colors, but DNA Painter gives you access to setting the colors in hexadecimal codes, so a bit over 16 million available colors...as if I can tell one from the other when the hues/shades are very close.

With a final caveat that the segment start and stop points we see from our testing and reporting companies are essentially estimates, not actual segment endpoints--you'll see different loci reported from different companies for the same people's DNA--even the free version of DNA Painter offers a nifty tool for organizing, sorting, grouping, and visualizing matches.

commented Jul 10, 2022 by Edison Williams G2G6 Pilot (447k points)

I took a closer look again at Leah Larkin's blog post linked above and saw that someone had posted, in the comments, a link to a palette of AncestryDNA Custom Group colors along with their hexadecimal color codes. However, there was one obvious error there--two colors with the same code--and, when I checked the color options at AncestryDNA, I realized that the palette had been created using the JPEG in the title of Leah's post, not the actual AncestryDNA color scheme.

So...having a rare 15 minutes free and being, er, sort of a stickler about these sorts of things, I built a PNG file showing the correct AncestryDNA color swatches and their corresponding hex codes. Just, you know, in case somebody somewhere might make some use of it someday...

https://countingchromosomes.com/downloads/reference/AncestryDNA-color-scheme.png

commented Jul 10, 2022 by Edison Williams G2G6 Pilot (447k points)

Answer 2 · 2022-07-10T17:39:41+0000

To expand on Harry's answer, here are a few things I'd try.

The 104 cM match with your maternal aunt can probably be useful as it doesn't appear that your parents shared any ancestors recently enough to cause such a significant match (they're 14C1R, so not likely to share any DNA). 104 cM is probably closer than 3C, so they would likely share 2nd GGPs. I only see three missing sets of potential CAs for them in your tree at that level so it's not likely that relationship is based on documented ancestors. There are a couple of common surnames, but they don't seem to share any recent common ancestors. You can use the "People who match both, or 1 of 2 kits" at GEDmatch on your two aunts and the results might help shed some light on the mystery. Those people who match both of them will have a good chance of being descendants of an ancestor of the unknown ancestor who should be on your mother's side rather than your father's. You can also use the "Shared matches" tab at Ancestry to see all those people who match both of your aunts.

Next, I'd upload your raw data to FTDNA. They have a "Not in common" tool so you can get a list of all the people who match your aunt but not you. You could also do this with your sister and your brother (when his results come in). Using Edison's handy equation I'd say that anyone who matches your aunt but doesn't match any of you three siblings is a good place to start looking (how's that for precision?).

Ideally you could find a 2C for your aunt who is a descendant of one of her paternal grandfather's siblings and have them tested. It looks like he had enough siblings that there should be a pretty good chance of finding a living descendant. No match there would be pretty solid evidence, though there's always the chance that the issue could be on the cousin's side rather than your aunt's. There's probably a reasonable chance that there's already someone out there who's tested that fits the bill. You'd find them in your matches or one of your siblings, most likely at Ancestry. You could also go to a 3C, in which case you could look for descendants of siblings of either John Jackson Tidwell (1835-1895) or Mary Elizabeth (Jones) Tidwell (1851-1917). Just keep in mind that there's a small chance that they might not have any matching DNA, as per the link Harry provided.

Finally, it's a long shot, but does your aunt have any significant component in her admixture that you or your sister don't? I' don't typically place much faith in the admixture results, but a glaring inconsistency can be a good clue. Best of luck, and let me know if you need any clarification or help using the tools.

Answer 3 · 2022-07-12T13:10:38+0000

Yeah, 630cM is pretty low, even for a half-aunt. Something's fishy.

When something like this pops up, what I want to see is the full picture - how well does everybody match everybody? You never know when some sort of fluke will strike. I'd especially like to see how well your sister matches your aunt, and your brother too, when his results come in. Both the average of the three results, and their spread, tells us more than just your test vs you aunt does by itself.

A full "matrix" of results would be (1) You vs Aunt - that's the 630cM; (2) You vs Sis; (3) You vs Cousin; (4) Sis vs Aunt; (5) Sis vs Cousin; and (6) Aunt vs Cousin.

I wouldn't say 104cM is "very low" at all. That likely makes the two aunts either 2C or 3C - they share a pair of either great-grandparents or gt-gt grandparents. If your parents are of no known relation, that tells you that your father's sister has a biological parent who is not one of your father's biological parents, and who is a relative of your mother. For example, your paternal aunt could be the product of an affair between your paternal grandmother and a 1st cousin of one of your maternal grandparents.

Keep in mind, however, that there are two ways for her to be a half-aunt. I mean that it could be your father who was the product of some affair, instead of your aunt. That being said, the fact that your paternal aunt appears to be related to your mother makes me think that she's the one who is a product of an extra-marital affair.

Do any other of your DNA matches on your father's side match you mother too? (You look at the Shared Match list, if you're not familiar with that indispensable feature yet.)

answered Jul 12, 2022 by Living Stanley G2G6 Mach 9 (92.3k points)

Thanks!

So, no "curve balls" regarding you vs your sister, or your aunt vs her daughter (hey - you never know, when this stuff happens!).

You're seeing the same basic thing with your sister's results, although she matches a little bit better. The averages for you & sis are 667cM & 298cM.

Assuming there's a generation between you and your aunt, that makes the case a little better that she's your father's half-sister, vs his 1st cousin. The average for a half-aunt is in the low 800s, and for a 1C1R is in the low 400s.

The result vs your cousin is the same story (which isn't surprising). A half-1st cousin averages in the low 400s, and 2C averages about 200cM.

The way to really tell is to identify your best matches that are on your paternal grandfather's side, and the same for your paternal grandmother's side. If your aunt doesn't match the people in one of those two sets of your matches, she's a half-sister of your father's, plain and simple.

For example, suppose you have 2nd cousins on your paternal grandfather's side in your DNA matches - grandchildren of your grandpa's brother. They're supposed to be 1C1R to your aunt, so if they don't show up on her DNA match list (and they would be pretty high matches), then grandpa isn't her father.

You could use 3rd cousins on that side also, but the closer the relation, the easier they will be to spot on your DNA list, and the less likely it will be that your aunt doesn't match someone because of chance. Any relative that is 2C1R or closer to her should be on her DNA match list with near certainty - although you might need to use the search feature to find them on her DNA match list.

BTW, I have two nephews (both are sons of the same brother) who have tested and may matches with them are 1494cM and 1607cM. My wife's match to her uncle is 1856cM. So those are the kind of numbers you'd expect for a full aunt.

Assuming we're right - that your aunt is a half-sibling to your father - roughly half of her DNA matches are for the relatives of her mystery parent (probably a mystery father). Your next challenge would be to use those matches to try to figure out what family her biological father came from.

commented Jul 12, 2022 by Living Stanley G2G6 Mach 9 (92.3k points)

Another possibility is that she's NOT a generation different from you. Every once in a while I see a case where the oldest daughter in a family has a child out of wedlock. The grandmother of the child is near the end of her child-bearing years, so the child is simply raised as if she is the grandmother's daughter, rather than granddaughter, and most people are none the wiser. The child is informally adopted by its maternal grandparents, and all the other kids refer to it as their sibling, even though it is their biological niece/nephew.

So if your aunt is the "baby of the family", it could be that she's really the daughter of one of your father's oldest sisters, if he has any. It's worth looking at the age difference between your "aunt" and your father. Your "aunt" could really be your 1st cousin.

commented Jul 12, 2022 by Living Stanley G2G6 Mach 9 (92.3k points)

Categories

How to find out about DNA surprises?

Please log in or register to add a comment.

Please log in or register to answer this question.

4 Answers

Please log in or register to add a comment.

Please log in or register to add a comment.

Please log in or register to add a comment.

Please log in or register to add a comment.

Related questions