PDA

View Full Version : Interpreting DNA data when comparing Ancestry DNA results with Living DNA -an anomaly



GG69
10-24-2020, 05:43 PM
Recently my Living DNA family matching showed a match at 1.57%(114cM) with a possible relative (rel A). The paper trail suggest she was a half second cousin and we shared a great grandparent (My grandfather's father was unknown and we we suspected it was this man). All the other information we have fitted this idea. This great grandfather also had another family abroad and the DNA of a relative (rel B) was available on Ancestry.co.uk. This relative uploaded it to Living DNA and eventually matches appeared. However the levels of matching were not as expected.
All three of us should have the same great grandparent and are all three generations below him
My match to this relative was only 0.36% (26cM).
Relative A's match to her was much higher at 3.37% (245cM).(In fact it was higher than my match to her.)
Now in theory if we are the same number of generations below him I would expect the matching values to be roughly similar but they are way out.
How accurate are these matching figures?
Could the anomaly be a result of comparing Living DNA results with Ancestry which may not use entirely the same markers.?
Alternatively should I be looking at one of his relatives as either way the three of us are related in some way?

pmokeefe
10-24-2020, 06:14 PM
According to this chart:
https://isogg.org/wiki/Autosomal_DNA_statistics#/media/File:Shared_cM_Project_v4.png
Half 2c
120 (average)
10-325 (min-max)

GG69
10-24-2020, 07:42 PM
Many Thanks pmokeefe. Very enlightening. The ranges for cM values are huge! I assume they are 2SD ranges. I suppose at the end of the day when you get big variations within possibly related people it comes down to looking at probabilities and fitting it in with other info one has relating to the family.
Trouble is people often look at numbers as absolute and don't realise the natural variation. (I was a Clinical Biochemist so am used to that!)

pmokeefe
10-24-2020, 09:35 PM
Many Thanks pmokeefe. Very enlightening. The ranges for cM values are huge! I assume they are 2SD ranges. I suppose at the end of the day when you get big variations within possibly related people it comes down to looking at probabilities and fitting it in with other info one has relating to the family.
Trouble is people often look at numbers as absolute and don't realise the natural variation. (I was a Clinical Biochemist so am used to that!)
Here's the page that was the source of the table, there are additional sources for statistics given there, including some from specific testing companies:
https://isogg.org/wiki/Autosomal_DNA_statistics

This document gives more details about the methodology of the table. It appears that the min and max are from actual data. But standard deviations are also given in the document. The values in the document are slightly different than the table accessible from ISOGG, maybe the sample is larger?
https://thegeneticgenealogist.com/wp-content/uploads/2020/03/Shared-cM-Project-Version-4.pdf

GG69
10-25-2020, 08:54 PM
Many Thanks

msmarjoribanks
10-27-2020, 01:30 AM
Helpful site for figuring relationships:

https://dnapainter.com/tools/sharedcmv4