Because the Y-DNA and MtDNA follow the direct paternal and direct maternal lines of the assumed family tree from the tested individual, any difference in the haplogroup from one generation to the next is an indication that the genetic relationship is broken.
A son should have the same Y-DNA as his genetic father, genetic paternal grandfather, genetic paternal great-grandfather and so on back in time. Similar for the mt-DNA and females, though males also carry mt-DNA but do not pass it on to children.
The scenarios to look for are determined by which DNA is different across the generation. As with anything, there is the possibility of an invalid test but in terms of the timescale we are looking at, Y-DNA and mt-DNA are immutable and do not change significantly.
The usual technique is to go back a few direct generations in the tree and then descend down another direct generational path of the same sex to find someone else who has tested for comparison. Direct male lines should have the same Y-DNA, direct female the same mt-DNA. This is where the larger database of FTDNA has an advantage as there are a greater number of possible DNA hits. It also requires a reasonable amount of family tree lineage to provide the documentary context.
If there are differences in the Y-DNA or mt-DNA between generations, the paper history is likely to be incorrect. The assumed (documented) parent(s) and the genetic parent(s) are different so you need to look at scenarios that fit that outcome. If both the Y-DNA and mt-DNA differ from other siblings in the family, a simple adoption from outside the immediate family may be the reason.
A father in a second or subsequent marriage will produce children with different mt-DNA to their older siblings, but as the father is common children will have the same Y-DNA. In close knit communities where marriages are seldom made outside the community, it could be possible for wife 1 and wife 2 to have the same mt-DNA and so produce children with the same mt-DNA and Y-DNA making it difficult to differentiate. This is one area where AuDNA is useful as an additional tool to separate the genetic contributions of wife 1 & 2.
If the Y-DNA is different it could be the wife had a child with a different partner or previous marriage, or a sister of the wife handing over a child to be brought up. The sister and the wife should share the same mt-DNA. Another possibility is a mother or aunt bringing up the child of a young or unmarried daughter/niece as their own, again not uncommon.
A difference in the mt-DNA could indicate the male of the household had a child from a previous relationship or marriage, or a recent relationship outside of marriage.
To work you need relatives to share their DNA to enable as many comparisons as possible to be made - at some stage someone has to be brave and be the first to take the tests and upload and encourage others in the family to join in too.
Although the Y & mt DNA database is smaller the information you get is accurate and can be definitive in stating if a relationship is valid. If the Y-DNA matches father to son, you can say that it is very likely they are related, remember another male in the vicinity with the same Y-DNA is still a possibility (a brother of the father or a paternal male cousin for instance). If there is no match then that father and son are not directly related genetically.
Sometimes this process can throw up numerous family contradictions too. Many well documented historical families have found that their robust family tree is slightly shakier after Y-DNA testing has confirmed that the heir-line to an estate is in no way directly related to previous generations due to a Y-DNA mismatch.
For surname changes, the FTDNA family name interest groups map surname against DNA haplogroup types. In my own case my surname is Jones and my Y-DNA is J2-L26. If I look at FTDNA there are very very few J2-L26 entries with Jones as a surname. However one of the highest occurrences is the surname Montgomery. This opens up a few possibilities.
A NPE sometime in the past. Probably the most likely.
A simple change of surname.
The Jones dataset is too small.
Did the son of a simple John from/of Montgomery take the Welsh patronymic form ap Ioan (John) becoming Jones when he moved across the border to Shropshire? It's likely the Scottish Montgomery Clan took 'de Montgomery' with them on their move to the Western Isles.
The DNA to surname mapping is also useful in finding geographic clusters of names with similar spellings, are they all related genetically?
@phil57
The LivingDNA test was an opportunity to get some data at a reasonable cost. A tie-in with Find My Past (3 months free full access)and provided an AuDNA file (approx 65k data points), a YDNA file with 288 SNP data points and a mtDNA with only 22 data points. The LivingDNA Y-DNA prediction did tally with YSEQ Cladefinder though even if it is a long way up the Y tree.
Inevitably, I see myself taking a Y700 in a few years time.
