au-DNA Relationships

+8 votes
280 views

Using au-DNA and GEDMatch setting of SNP=500 and cM=1, I have made 7 matches in multiple segments with my immigrant.  He is a 5th great grandfather, and all 7 matches show a solid paper-trails to our immigrant.  Thes low settings were recommended by Andreus West a DNA expert in Germany.  I have identified another ancestor with 3 individuals who match my 7 but with no paper-trail. Some have identified this new match a son of the immigrant.  Is this enough matches to say that there is a close relationship, and they have to be father-son, nephew- uncle, brother or close cousins.

in The Tree House by Anonymous Barnes G2G6 Mach 3 (32.7k points)
Thanks Hugh. BTW, I was born in Germany but I live in Singapore ;-)

Maybe you want to show everyone all the matching segments and kit numbers (either in a tabulated format here on WT or link to an uploaded image of a screenshot of your spreadsheet).

Then everyone can take a look at your work and comment on it. Otherwise the discussion with the mostly fruitless and theoretical.
Revealing kit number in most cases displays actual name and would potentially be an invasion of privacy.   For this reason, I must decline.
Hugh, fully understand. How about disguising them with a running number, like kit1, kit2, kit3 etc.? You could still post the matches that way and if anyone would like to help you they could still get in touch with you personally.
I will add privatized spreadsheets later.

2 Answers

+3 votes
 
Best answer

Guys, just because things were done in a certain way in the past doesn't mean that a new way is wrong. Or do you still belief that the Earth is flat?

500 SNP's are 500 unique markers in a row. If that is stretched out over 1cM or 5cM is no difference. The only difference is that recombinations are suppose to happen 5 more times (or are 5 times more likely). Check the definition for centiMorgan, it's not centimeter.

What are the chances that 500 markers are identical by chance? Can one of the Math buffs here on the list come up with an estimation?

I'm aware of all the old studies and papers written about this and yet, if you apply triangulation and you find many DNA kits that triangulate into this group plus there is a paper trail to a CA with several MRCA then it's proven for me.

Please keep in mind that based on my work in this area it's common that you have matches with a minimum of 700 SNP's / 7cM in this TG, you also have some matches that are above 500 SNP's / 5cM in the same TG and yes, some DNA kits have several (usually 2) broken up segments within the overall boundary of the TG.

The key is to apply triangulation and have at least one DNA cousin in the TG with over 700 SNP/7cM.

by Andreas West G2G6 Mach 6 (60.6k points)
selected by Anonymous Barnes
+8 votes
PLEASE read:

http://www.wikitree.com/g2g/199470/are-you-using-audna-segments-which-are-less-than-7-cm?show=245672#a245672

Jim Bartlett understands auDNA matching better than almost anyone.

Lowering the matching threshold below the recommended level is a recipe for "confirming with DNA" genealogical mistakes.  Millions of people living today have a birth certificate (solid paper trail) stating who their father is and in reality he is not their biological father.

Most sincerely, Peter
by Peter Roberts G2G6 Pilot (559k points)
I'm not so sure you can't get sound matches from multiple matches less than 7 cM but 1 cM  is stretching it beyond the breaking point.  The best test would be to take someone with a match in a line which doesn't include the person in question and see what it shows when a match is tried.

See:

https://segmentology.org/2015/09/30/small-segments-and-triangulation/

It is only when both people are matching with phased auDNA that you can use small segments with confidence.  http://dnamatches.blogspot.com/2012/11/autosomal-dna-testing-phasing.html

(side note: AncestryDNA uses pseudo phasing which is not the same.  Plus AncestryDNA does not provide tools to view shared segments.)

Related questions

+12 votes
2 answers
+11 votes
3 answers
+4 votes
1 answer
+10 votes
1 answer
+8 votes
3 answers
+4 votes
1 answer
+14 votes
6 answers

WikiTree  ~  About  ~  Help Help  ~  Search Person Search  ~  Surname:

disclaimer - terms - copyright

...