New and Updated Articles

Can I trust this DNA stuff?

This article is for those of you who have reservations about DNA and its place in Genealogy. You may not trust it; you may be sceptical about the supposed science behind it, or have concerns about privacy. You may even think it is just a scheme for companies selling these DNA tests to make a fast buck of naive customers. But at the end of this article, I hope that some, though perhaps not all of these beliefs will be challenged and that you might actually want a DNA test yourself.

So, let’s deal with the following points in turn. Feel free to read the whole article, or to click on a link to skip to the part, or parts, that interest you.

Dodgy science?

A good genealogist deals in facts. When you're researching a family tree, you may be looking for say, a baptism record for a Joe Bloggs. You might find a website that claims a Joe Bloggs was baptised in London, England in 1822. You're pleased you found this information but what you really want is to see the evidence - the source of this claim - because it might turn out that he wasn't born in 1822 at all, but rather 1922. In genealogy you need to see the facts.

Likewise, good science is built entirely upon facts, and to be of value they have to be useful facts. So is this DNA genealogy based on science, and is it useful?

The technology used in Paternal Y Chromosome tests for genealogy came off the back of work done by professor Alec Jeffreys at the University of Leicester, England and was announced in 1985. Professor Jeffreys has since received a knighthood and is now Sir Alec Jeffreys. Sir Alec’s work was based upon other DNA discovery over the years which started in 1869 when Swiss scientist Frederick Meisher isolated DNA from puss samples on used bandages. You can read all about it here on wikipedia.

The subsequent tests were developed in academic laboratories in the 1990s. One of these academic laboratories was at Oxford University in England. About the turn of the century these tests were first put to commercial use and developed as products sold to the public. One of the early companies was Oxford Ancestry who are still around today, a spin off from the university. FTDNA and Ancestry were founded and over the last decade became the market leaders.

Alongside the growth of the public market, universities and other institutes have continued using these tests to carry out archaeology and other science projects, including purely genealogical ones.

One such example is The Genographic Project, a not-for-profit venture by The National Geographic Society and IBM with most of the gain spent on cultural preservation projects nominated by indigenous communities. More than half a million people have partisipated. These types of efforts, both in the public and private sectors have driven the technology and analysis forward, and importantly has resulted in larger, more complete reference results from static populations.

The level of adoption is a really important factor. Back in 2000 when my father first took a test with Oxford Ancestors, the results were compared against their database at the time, and there weren’t that many results to compare against. We had one possible match in the US and little in the way of analysis from the company selling us the tests.

There was also very little information in books or on the internet that could be accessed to understanding your results, and there was no free database of results that you could join, no matter which company you tested with. Today, all these things are available.

Now, more than a decade later, when you take a test you get to compare your results with hundreds of thousands of others and can access a large amount of information and expertise to help you understand your results and get value out of them (more on value later on)

In comparing your data with other results, it is important to understand where these results come from. The primary purchasers of DNA tests by the public is from American countries. Second, probably is the UK and Australia, followed by countries in Europe. You don’t get many people from less wealthy nations taking tests. However, universities do a good job of collecting samples from more static populations like these as the data is important in understanding the overall picture of where we all come from.

All results are analysed and put into categories called Haplogroups, Each defined by different marker numbers. Each Haplogroup has a different geographic origin and occure at higher frequencies in different places. By far the most common is Group R which makes up about 70% of the population of Britain, and is almost 99% concentration in North West Ireland.

There are benefits and drawbacks of belonging to a common and diverse Haplogroup such as group R. The obvious benefit is that there are lots of results to compare against of people who may be related to you. The drawback is that you have to work out which ones are genuine matches that you are related to and which are coincidences.

Not too hard if your willing to learn how, or if you have someone who will do the analysis for you.

If you belong to a rarer group, like J (as I am) then you know that close matches are related and are highly unlikely to be coincidences. But the drawback is that there are less results to compare against. For every forty or so tests that someone in the world takes, only one will belong to the same haplogroup as me.

So who decides these Haplogroups?

There are a couple of different bodies, some are early adopters of new evidence and others more conservative. The most recognized body is ISOGG - The International Society of Genetic Genealogy who occupy a good middle ground, being a little bit conservative, but not behind the times.

Companies selling tests like FTDNA and Ancestry are typically very conservative, this is because they don’t want to get egg on their faces in front of their customers or to upset their customers by classifying a customer one way and then changing their mind later when those science findings don’t work out.

Chances are, if you accept FTDNA or Ancestry's classification of your haplogroup then it will be correct, because it’s based on science that has been replicated and verified by many studies and so on. If you are adventurous then you can read the more cutting-edge science discoveries in the area of genetic genealogy and further refine your haplogroup.

For example, your haplogroup might be J, and FTDNA might tell you that the subclade is J2a but the most recent work might indicate that your subclade is likely J2a1d. I think the point though is that it is a scientific process and the DNA test companies are conservative and slow to accept new findings, so you can trust them in this regard.

Because paternal Y chromosome tests have been around for a while, the science has progressed to a point where the haplogroups are well understood and quite detailed already. This is not the case so much with new tests such as Autosomal tests (more on this further down).

The point is that any new technology is adopted by a minority who pay the initial high cost of having that new technology. Over time the costs of production comes down, and with it so does the purchase price as the same or better version of the product is sold to new people who are subsequent adopters of the technology.

The same is true with DNA genealogy tests. So new types of tests are more expensive with less benefit, and carry the risk of being dropped (like Betamax bring dropped over VHS). So it is better to go with established tests like the Paternal Y chromosome test, rather than the newer autosomal tests - the science here is early and subject to revision.

This takes us nicely onto the next topic - different types of tests.

Differences in types of DNA tests

paternal Y chromosome tests provide data on a type of marker known as an STR (Short Tandem Repeats). These mutate fairly slowly. You might expect only one marker out of 37 to mutate in 300 years. This means that when you look at two men sharing a common male ancestor 300 years ago, they might be a 36/37 match, and the one mutation actually helps you to estimate how long ago they shared their common ancestor. If they had 2 or 3 mutations then you would be able to estimate that their common ancestor was further back in time, perhaps 6- or 7-hundered years ago. Im speaking generally here, as different markers are known to mutate at different rates. The point is that these STR markers continue to match close to 100% up and down many generations, including cousins relationships if they are in the same male line and any small number of mutations helps you to know how far back two people shared a common ancestor.

From a genealogical standpoint therefore, Y DNA tests are by far the most useful. This is because the Y chromosome is passed down unchanged from father-to-son down the generations, and because surnames have also been passed down in the same way down the generations, the two pieces of data (DNA and surname) can be used to make useful conclusions about a person’s ancestry. The drawback is that these tests can only be taken by a male and therefore only sold to a male, or a female who has a male relative (more on that lower down).

Another type of test is the mitochondria test. This is similar in that it is passed down from mother to child (male and female) and so your mitochondria in your cells today is almost identical to your mother's mother's mother and so on centuries, even millennia ago. But, it is not as useful as the paternal test because surnames aren’t handed down from mother to child, but rather from father to child.

Then there are autosomal tests. These are relatively new, and are being promoted heavily by all the genealogy DNA companies. Most of the new autosomnal tests use 700,000 markers or more, known as SNPs (Single-Nucleotide Polymorphism). SNPs mutate very slowly, slower than STRs, and once mutated are stable and are an excellent signature of a persons haplogroup, able to inform regarding a person's deeper ancestry (thousands, to tens of thousands of years ago).

But, because they mutate so slowly, they are less useful for gaining information on a persons genealogy in recent times (around the last 1000 years).

Although around 700,000 markers, this is still looking at a relatively small portion of your overall DNA, and it is spread out across 44 of your 46 pairs of chromosomes. The technology that produce these results, provide data on all your 46 chromosomes, but the test companies remove the data on the X and Y chromosomes. They do this because they want to sell you that information seperately as Y chromsome tests and mitochondria tests.

The flaw here is that most of these chromosomes are a mixture of the DNA from your two parents, and their chromosomes were likewise a mixture of their parents and so on. So, a marker on an autosomal test could have been inherited by you from anyone in your tree and you don’t know who. It could come from your great great great great great grandfather, but it’s equally possible that it comes from your great great grandmothers father's father's mother, or any other combination of your ancestor relatives.

They therefore provide accurate information on recent relationships, so people related to you in your generation (siblings) and one either side (your parents, and your children) will share about 50% of your autosomnal results. But chances are that you know who these relatives are already.

With each new generation the amount of shared DNA is halved. So, you will share about 25% of your DNA in common with your grandparents, and 12.5% with your great grandparents and so on.

ISOGG provide some statistics on autosomal DNA results, showing the various degrees of relationship, including cousin relationships.

This halving of shared DNA in each subsequent generation makes it difficult to draw any useful genealogical data from this type of test. It might tell you that one of your relatives at some point came from a distant place, but so what? I mean, when you do the math you can see that you have an enormous number of relatives that have contributed to your own DNA, go back just ten generations above you and you have 2,046 ancestors:

2 Parents

4 Grandparents

8 Great Grandparents

16 Great Great Grandparents

32 Great Great Great Grandparents

64 Great Great Great Great Grandparents

128 Great Great Great Great Great Grandparents

256 Great Great Great Great Great Great Grandparents

512 Great Great Great Great Great Great Great Grandparents

1024 Great Great Great Great Great Great Great Great Grandparents

And all of these have contributed to your DNA (with the exception of your Y-chromsome and mitocondria dna, which are restricted to the direct paternal and maternal lines respectively) So it’s no surprise that we each have ancestors from a diverse geographic background. This is the same reason why if you look hard enough just about everybody is descended from King this and Queen that - just not in the direct line.

If you look at the example below you can see that the green man at the bottom that represents you or me is getting DNA from all his ancestors in red, and this diagram is only looking at four ancestor generations - that's already 30 people that are contributing toward your DNA, and you don't know which of these people gave you which bit of DNA.

All ancestors

Compare that with the following diagram which represents where the green man's Y-chromosome is coming from:

Male line

As you can see the green man knows exactly who gave him his Y-chromosome, and every male in this line (father, son, brothers, uncles, cousins) will all have the same Y-chromosome, making it a reliable and useful piece of data for genealogy.

These autosomnal tests are getting better, and may be valuable in the future once databases contain larger reference populations, but right now I dont feel I could reccomend them to anyone.

So why do companies sell these mitochondria and autosomal tests?

What Companies like about these tests are that they can be sold to anyone. Male or Female. That is clearly a big motivation. In addition though, you have to bear in mind that these tests may become more useful over time (though i doubt they will even be as useful as a Paternal Y chromosome test) and so if you are a company selling these tests, you do have to start somewhere, and with all new technology there are early adopters who are the foundation for the masses of customers later.

Another factor is that if you have a competitor who starts selling an autosomal product, you cannot afford not to offer a comparable product. If you don’t and it takes off and becomes useful years down the line, you have already missed the boat because your competitor has a large database of results to base findings on and you do not.

These are the reasons why FTDNA and Ancestry have recently offered these autosomal tests, and each probably did it in response to companies like 23andme (primarily a personal genomics company focusing on genetic health information) that tell their customers a bit about their autosomal genetics.

All this doesn’t mean you should buy an autosomal test mind you, or a mitochondrial one for that matter, and that brings us onto our next topic.

A money making scheme?

As we have established earlier in this article, Paternal Y Chromosome tests are the most useful for genealogy, but they can only be sold to men. Hence the companies desire to come up with tests that can be sold to anybody, which doubles their potential customer base.

If you do want a Paternal Y chromosome test then there are different types. Variations exist between companies and you can buy different numbers of markers - the more markers the more they cost. You can read an article about these variations here that will help you get what you want for the least amount of money.

Mitochondria tests tell you something interesting, but from a genealogical point of view, I don’t think they are of much value and I wouldn’t recommend buying one.

At the moment The autosomal tests are, in my view of almost no value at all genealogically. I think these tests are mainly about making money for the companies.

How do you know that the results aren’t made up?

This is established science, now almost two decades old, backed up by thousands of published science studies from reputable institutes. There are science journals dedicated entirely to this area of research and you can take a degree on the subject at many universities.

That said, if you still have some doubts it is possible to put the science to the test.

I've done this myself. If you look at the first two results in the Langton DNA Project.

you will see the results of a father and son, David Langton and Joel Langton. David was tested about 2000 by Oxford Ancestors, and Joel was tested in 2012 by FTDNA. You can see that the two results are identical, as they should be, because Joel inherited his Y chromosome unchanged from his father.

Privacy issues?

Some people worry about their DNA falling into the wrong hands. In theory the DNA you provide via a mouth swab, could be enough to make a clone of you, or to identify some of your health traits.

However, there are a few facts to consider. Firstly, why would anyone want to make a clone of you? Have you seen that Star Trek film, Nemesis? In the film the Romulans took a sample of Captain Piccard’s DNA and cloned a younger version of him with the aim of replacing Piccard with a version that they had control over. Well, there is a motive at least, but unless you hold some vital position in society no one is likely to want to clone you. Also, as shown in the film the plan didn’t work because the clone was younger and even with the supposed technology of the 23rd century they couldn’t accelerate his growth enough without killing him.

Secondly, it still costs millions of dollars to clone anything and so no-one is going to fund making a clone of you, and it will probably always will be very expensive. It is also illegal in much of the world and where it isnt, there are still lots of prohibitions in place that would make it difficult for would-be-cloners to carry out reproductive-cloning.

Thirdly, it is worth bearing in mind that if anyone wanted your DNA that bad they could easily take it as you leave DNA on every cup you drink from, and leave stray hairs wherever you sit and these samples are all plenty big enough to extract your DNA from. Though DNA theft is illegal in many parts of the world, including the UK since 2004 (see, Human Tissue Act), so its probably not worth worrying about.

Some people worry about the results themselves; what if someone got hold of them?

These results that companies like FTDNA or Ancestry provide are on a few dozen markers of your DNA, and represent a tiny tiny miniscule portion of your overall DNA. There is no medical information in them and you literally cannot do anything with them other than identify who you may be related to. A genealogical DNA test is basically the same as a DNA paternity tests.

You know those tests that they do on shows like Jerry Springer? Well, they are basically the same. Your results cannot be used for anything else, beyond that, there is nothing sensitive about them whatsoever. So, you should be more concerned with shredding your bank statements before putting them in the bin (you do shred, right?), than worrying about you’re DNA results falling into the wrong hands.

Some people might associate DNA tests with criminals that are DNA tested in murder enquiries and so on, but these forensic science tests are different in that they look at a much larger portion of a person's DNA. They are not the same as a genealogical DNA test, and your results wont and cant be used to compare against any forensic DNA held on any criminal registers.

The law also enures that you retain control over your own DNA. If a DNA company that you tested with were to be sold for example, you would have the right to withdraw your results from the database. The new company would not see sight of your results. The results are not automatically transferred to the new company, instead you must opt in if you want the results transfered.

DNA companies have strict privacy policies, and most of these companies (including both FTDNA and Ancestry) are also compliant with something called the Safe Harbor Privacy Statement which gives additional protection for EU citizens. Here are FTDNA and Ancestry's privacy policies:

If you still aren’t comfortable with sharing your results with others then do bear in mind that you can opt not to share them with anyone when you buy the product from FTDNA or Ancestry. And most DNA projects, including the Langton DNA Project, allow you to display your results anonymously. This means that your results will be public but it would be shown as belonging to a Mr Anonymous Langton. No one will be able to identify you personally and we don’t share any of your details with anyone else without your express permission.

What about non-paternal events?

You sometimes read in the media that the results of DNA paternity tests show that a high percentage of babies born are secretly fathered by someone other than the father who is on the birth certificate, with figures between 10 and 30 percent often quoted.

However, the figures quoted are based upon paternity tests taken, and yet these are not taken from an unbiased sample of the general population. People usually have paternity tests because they already have doubts about the paternity of a child, and these people actually make up only a tiny fraction of the general population. So of this tiny portion of the overall population only 10-30% are found to have had a non-paternal event. In other words, it is really much much rarer than the media would have you believe, but if they ran with the headline "1 in 10,000 children are fathered by an imposter" rather than "1 in 10 children..." you get the picture, they wouldn’t sell many papers.

I know of one person who suspected a non-paternal event but they went ahead and had a test anyway and his results showed that he belonged to the family that we thought he should and matched very very closely. The results were conclusive, he definitely did not have a non-paternal event.

Even in a scenario where a non-paternal event has occurred, the chances are that the event happened many generations ago. Your DNA result may give you clues as to your real genetic heritage in an unexpected direction but you are still who you always were. The non-paternal event would have already happened before you were born even if you knew nothing about it. Better to know it and discover facts about your past, than think false things about your ancestry.

Your identity is still yours, you are still a mixture of genetics and upbringing and individual choice. Your name remains the same, and the person you are remains the same.

And, as I have said shown, it is really very rare, so you dont need to worry about it.

Value for money

Since these tests became available in the 90s, the cost has come down significantly as more people have used them. Back in 2000 my dad paid more for a 12 marker test than I paid for my 67 marker test in 2010.

In addition only one male member of your family need take a test, as your results will all be the same. So, for example, if you have two brothers then you could split the cost between the three of you and only one of you do the cheek swab, safe in the knowledge that the result will be the same regardless of which one of you takes the test.

OK, so they aren’t prohibitively expensive, but where is the value? These tests cost money and perhaps that money is better spent on some documents from the archive or a new shed?

The least that you will get from a Y chromosome DNA test is a haplogroup, subclade and a list of matches. This will tell you some of your deeper ancestry, in my case I know that my ancestor about 4000 years ago was living in the area around Mesopotamia, and with some analysis we discovered that subclade was almost exclusively of Jewish origin (this was an unexpected surprise). Your list of matches may contain people with the same surname as you in which case you are related, and it may contain some names different from you. In most cases these other names are related, but only very distantly, with your lines going their seperate ways before surnames were taken on from about the time of the Norman Conquest on Britian. You will be able to tell from the results just how closely related you are and will be able to roughly calculate how long ago you shared a common ancestor.

Of course, your chances of a match are higher if other people with your surname from around the world, have already taken a test themselves. You can check if there is already a project running for your surname at FTDNA and Ysearch to see before you think about buying one.

There is more potential value in a DNA test if you have a place-name surname like Langton. This is because it is possible to match a DNA profile with a specific family from a specific location where their name was taken circa 1100AD.

For example, if someone with the name has remained in the vicinity of the place-name, in the case of Langton this would mean someone living in, or close by, a village called Langton and with a tree showing they have been there for some time, then you can test that person and know with a large degree of certainty that their family took their name from that village 900 years ago. Now that you have their DNA profile, anyone who takes a test and matches that profile, will also have come from the same family, and their ancestors from the same village.

You can also do this with non-static populations, but it’s harder because you have to be able to reliably trace their family tree back to the village of name origin. The Langtons of Langton-by-Spilsby have held the estate since at least 800 years ago for example, so we know that anyone who matches their DNA profile is from the same village originally. Though this is a rare example in the UK.

If we look at what I have got from my test, I have found that I came from Langton-by-Horncastle between about 1066 and 1200. I had to work this out myself, and I couldn’t have done that had there not been other test results to compare against, but millions of people have taken Y-tests now, so you often get alot of good matches now. The matches that we had were enough to show that we came from Langton by Horncastle in Lincolnshire.

I don’t match any Langtons yet, except my father, but as more Langtons take tests as time goes by, eventually another Langton will match me. This will tell me that we are related and give me clues as to my family tree.

Tradtional genealogical methods have taken my family tree back to Newbury, Berkshire, England in 1679 and despite a lot of effort I have not been able to trace back further.

There are several Langton families in the early 1600s in neighbouring counties but I cant find any records to show which my family descends from. In this case, the records have failed me here, and DNA is my saviour. Without DNA I would not know where I came from before 1679. I don’t know how I got from Langton by Horncastle in Lincolnshire to Newbury, in Berkshire, but I DNA has helped rule out a couple of the possibilites as we know they belong to other families with different DNA from me.

In time, I expect that another Langton will take a DNA test, match me, and have a tree that goes back somewhere close to Newbury, via one of these other Langtons in neighbouring counties, and i will know that I join that group somehow.

Only time will tell, but this is the reason why alot of these projects exist, people want to connect with distant relatives and discover new things about their tree, and DNA is one way that you can do that.

It is also worth bearing in mind that I am in the minority, having no Langton matches yet. Most people in the Langton DNA Project do match other Langtons, our biggest group having eight members that share the same DNA profile and are related. Many have never met one another and some live on different continents. Some have since been able to work out how their trees connect, having been able to focus on that area, knowing that there is a connection somewhere.

So you get the results and the matches and some value comes straight away from your DNA test, but to get the full value out of it, you need to be prepared to do some analysis yourself (or know someone who will do this for you), same as you would with any traditional genealogical research effort. You can buy or borrow books about DNA and learn the characteristics of different markers from the results.

Of course, if you are a Langton, then you get the benefit of all our experience and analysis for free, ensuring that you get a lot more out of your results than you would otherwise.


Instead of seeing DNA as an opposing force in the world of genealogy, or as something that doesn’t even belong in the world of genealogy, try to see it as a tool in your genealogy belt (at least the Paternal Y chromosome tests), not replacing the good work of careful research spent in the archives - or as is the case increasingly nowadays, on the internet - rather see DNA as a different option for finding out facts about your past, that can help inform your more traditional methods, or supplement the data that your traditional efforts provide.

If you still aren't 100 percent sure you trust it then buy a cheap 12 marker test. It's better than nothing and although you often need more markers to get a conclusive match (particularly if you belong to a common haplogroup) it will tell you something, and if you are lucky enough to belong to a rare haplogroup, then it may be enough to get a conclusive result right away.

Most DNA companies (certainly the big two, FTDNA and Ancestry) let you upgrade to more markers at a later date if you want to.