DNA Analysis

spurious wrote: »

FTDNA has a lot of Irish (in Ireland) people mainly because it works out the cheapest.

What I always suggest is if people have tested at either AncestryDNA or 23andme that they create a FTDNA account and then transfer in their test results form these companies. You can do this without providing an additional sample for testing.

Of course subsequently if individual is male they can always request a upgrade to a basic Y-DNA str test (which will require a new sample). It's a fairly cheap way to get into their database.

Well, how are the results different?

pinkypinky wrote: »

Definitely am repeating myself but the ethnicity guide on Ancestry partially comes from placenames entered on trees belonging to your matches, so it may be accurate, but it's not coming from the DNA itself.

That has always been my understanding of the ethnicity section of DNA results ie its based on matches with available trees.
Have 3 tests on Ancestry and all have what look like accurate regions but then again all 3 are 100% Irish back to 1800 and most likely long before that.
All their ancestors were/are from very localised areas in specific counties so the tree matches reflect this.

Have one group in my own DNA match list with 200 plus likely members which almost all show as less than 20% Irish ethnicity but think that its more to do with the fact that all available trees apart from my own show not a single ancestor from Ireland which skews the result.

Useful recent presentation on BigY and Y-DNA testing in general here:

I have a 28cm match on living dna who is also signed up to ancestry but we don't appear as matches on ancestry?

As for the ethnicity estimate with over 200 years of direct ancestors coming from Cork and kerry it does give me 98% cork and kerry and an odd 2% german.

Nqp15hhu wrote: »

LivingDNA doesn’t go down to that resolution for Ireland

But its living dna that shows the match while ancestry doesn't, its actually 26.88 cm estimating at 3rd to 5th cousin.

Am I correct in presuming that all of these dna testing kits come from the US? I was looking to give one as a Mother’s Day present but I may have left it too late.

pinkypinky wrote: »

I think you would have to provide a new DNA sample for them to do the Y test but would really not advise Y12. You won't get anything useful from that - the minimum recommended is Y37 and even that is fairly unhelpful. As an indicator, I've got 3 Y37 tests on family members and have yet to find a single person in any of them with the same surname or make any concrete connection to anyone who is a match. There are not enough Irish people in the database yet.

I did a Y37 dna test with FTDNA and although I have matches, there are none with my surname and my closest match predicts our most recent common ancestor as 24 generations ago. This test might benefit me in the future but for the moment it appears to have been a waste of money.

Interestingly they were not able / willing to connect the Ydna test with my autosomal dna test, even though I had the former done as an ugrade to the latter.

bluezulu49 wrote: »

I did a Y37 dna test with FTDNA and although I have matches, there are none with my surname and my closest match predicts our most recent common ancestor as 24 generations ago. This test might benefit me in the future but for the moment it appears to have been a waste of money.

Interestingly they were not able / willing to connect the Ydna test with my autosomal dna test, even though I had the former done as an ugrade to the latter.

It can depend on your Y haplogroup, of course. The majority of Irish men fall under groups with large numbers due to a growth of the ruling elite and hereditary rules from back in the day.
I fall under haplogroup I and have about six Y matches. Most have a Scottish equivalent of my surname and points to either plantation or right before plantation ancestry on the male side.

so useful table:

Relationship Average % DNA Shared Range [1]
Identical Twin 100% N/A
Parent / Child/ Full Sibling 50% Varies by specific relationship
Grandparent / Grandchild / Aunt / Uncle /Niece / Nephew /Half Sibling 25% Varies by specific relationship
1st Cousin 12.5% 7.31% - 13.8%
1st Cousin once removed 6.25% 3.3% - 8.51%
2nd Cousin 3.13% 2.85% - 5.04%
2nd Cousin once removed 1.5% 0.57% - 2.54%
3rd Cousin 0.78% 0.3% - 2.0%
4th Cousin 0.20% 0.07% - 0.5%
5th Cousin 0.05% Variable
6th Cousin 0.01% Variable

So when you get down to second cousin etc.

So once ye down at '2nd cousin once removed' it's quite a broad range, but ye looking at tops of 2.5% shared DNA.

This is because the process of Meiosis that produces gametes (Egg's/Sperm) results in genetic recombination, basically your 46 Chromosomes (other than your Y-Chromosome if you're male) do not fully match chromosomes seen in your parents.

Instead each Chromosome is an amalgam of two of your parents chromosomes.



When it comes to the pair of sex chromosomes (X and Y) this is more complicated. The Y-Chromosome basically can't recombine with X (outside of very small regions on edge), as that would risk the loss of genes that encode for genetic maleness. As a result a daughter will basically inherit an unchanged X chromosome from her father, but the X she gets from her mother can be made up of variable mixture of mother's two X's



In case of a son, the Y-Chromosome is basically identical to father's (except if new SNP mutations arise in utero) but the X he has is mix of both his mothers. As a result two brothers could have quite varaible shared DNA on their X.



As a result you will have people who can be proven to be say 3rd cousins via standard genealogy but with whom you share variable amounts of DNA with. It's completely posisble to be genealogically related to someone but not share any DNA (as the process of recombination has filtered out over generations what would be shared). This is why autosomal tests (23andme/Ancestry etc.) are only really useful finding matches with time of divergence of a maximum of 200 years.

In comparison if you had two males with proven written genealogy showing they were 8th cousins (on male line with same surname), they should match on Y-Chromosome even though they share 0% autosomal connection (in all probability given distant cousin marriage they probably share more recent connections on autosomal DNA from lines other than their direct paternal lineage)

paddysdream wrote: »

Presume this is the correct thread for this.

Has anyone ever looked at how much DNA they seem to have from each of the different sides of their family lines ?
Know this is perhaps not a scientific method of doing things but my own matches seem to heavily weigh to a couple of particular lines regarding the amount shared with relatives of equal distance.
Discount number of matches as this can be biased if a family was large or small and in most cases a family line with a lot of US emigration will usually have more people tested on average.

This came to mind recently when a new match appeared on Ancestry for a 2nd cousin once removed .They happen to be on my fathers paternal side ,on which I have very few matches for that are placeable in my tree .

With this person (female)I share 39cM whilst with another person (male ) exact same relationship on his mothers side I share 185cM. Both people in question are grandchildren of 1st cousins of my father ,the first relationship through a sister to his own father (my grandfather) and the second through a brother to his mother (my grandmother).
Does the sex of the people involved have any bearing on all this ?

Looking at my matches some family sides show much stronger matches than others with known 3rd cousins varying from 195cM back to 23cM .
Have discounted being doubly related to people in this as in many of the cases they are either in USA or Australia and their tree indicates no connection to Ireland or in particular any area or name that might raise this possibility.

Again not too scientific but when you hear "well he takes after his grandfather/grandmothers side "or "he's a real Murphy/Byrne/Doyle/Smith etc etc " wonder if that follows through to shared DNA amounts.

I'm getting a weird result at gedmatch when I use their feature to see what mutual matches I share with a person.
So I plug in my id and the other relative who gedmatch estimate to be 4.7 generations out. She matches my Uncle, but not my first cousin and second cousin who are related to me via my grandmother (Uncle's mother). But those two cousins then show up as close as you'd expect with my Uncle.
I suspect this may be an issue with gedmatch's algorithm/variances between testing companies raw data.

Also does anyone know what sharing one very large segment with a person may indicate? I have a new match at 23andme where I have one shared segment of 0.81%, I think I read in the past that this is called runs of homozygosity and may be signs of pedigree collapse.

My mother is from Clare, she got 100% 'Irish' in Ancestry (with various North Munster "genetic communities")

My late father shows up as:

• Ireland: 89% ("Tyrone, Derry, Antrim" + "South Derry, East Antrim" communities)
• 9% Scottish
• 2% Welsh

I'm from Galway and show up as 97% "Ireland" (various North Munster communities) and 3% Scottish.

For context my Dad was from Athlone, his father was from Belfast (with a "Liverpool Irish" mother) and his mother from East Galway/South Roscommon (whose mother's parents were from Cork/Kilkenny)

As these tests are really only good back to 200 years ye only really seeing a snapshot of general Irish population in the recent past. eg. their 'Ireland component' is reflective of general Irish population over the last 200 years.

https://whoareyoumadeof.com/blog/what-are-the-pros-and-cons-of-genetic-testing-for-ancestry/#:~:text=%20What%20are%20the%20Pros%20and%20Cons%20of,post%20helped%20you%20understand%20a%20little...%20More%20