View Full Version : Misinterpretation of Halopgroup and SNP data

09-17-2014, 12:37 PM
It appears that Halopgroup and SNP data is being misused to identify family relationships in certain Surname Groups without reference to the caveats and assumptions used in these predictive models. At bes SNP data can give an indication as to the DNA journey. However when you take the unpredictable nature of DNA mutations and the potential of universal reversibility of both STR and SNP signatures rigourous scientiffic method needs to be applied when even analysing family lineage using Halopgroup / SNP indicators alone alone.

DF19 for instance only predicts potential North Germany / Danish origin about 2800 years ago. Plus or minus means little to nothing as mutations in one direction does not fit with the unpredictable nature of biological mutations and the error rates in samples. DNA is not a perfect recording devise. If it was we would all be clones.

Beware of data misuse

09-18-2014, 10:00 PM
SNP signatures are not reversible. You are right, DNA is not a perfect recording device. But it is almost perfect and much more reliable than a paper trail. And there's much less chaos in the mutation occurrences than you would assume. We can expect a certain amount of mutations per 500 years, and we know some STRs mutate much faster than others. Data misuse (denial, actually) would be a person using a paper trail as an attempt to trump a DNA non-match. It screams:

"here is a person who will not accept reality, who refuses to believe that a NPE has caused them to be a non-match with the family they want to match up to."

Genealogy can be a tough hobby until emotion and desire are removed from the equation. And there's nothing wrong with a NPE - I would guess every family has one in it somewhere. It can be hard to accept the DNA results at times, when these results aren't what we expect. But does it really change who we are? I think not. The toughest part of being an administrator in a DNA project is to be the bearer of bad news and confirm a non-match where a match was assumed. We (admins) don't do this to break hearts or destroy dreams. Instead, we hope to point these folks in a new direction and help them solve a new mystery. What good is it to lie to someone and tell them what they want to hear? If someone is willing to pay the money and swab the cheek, and they join a surname project, they deserve honesty. Why would they swab that cheek in the first place if they didn't want to know the results?? But be careful what you are saying about SNPs. If two people are supposed to share an ancestor 600 years ago, but one tests positive for a 3,000 year old SNP (for the sake of argument, let's say the SNP was DF19) and other tests negative for the same SNP as tested through the same company and lab, then the answer can only be that they do not share that ancestor outside of a NPE.

I hope this helps :)