A slightly modified version of this article appeared in Family Tree magazine in June 2013 (See after the text)
Testing our ties – Genealogical and genetic ancestry – What’s the difference?
Economist and historian Stephen Lewis puts our roots under the microscope to discover a little more about how we inherit some genes and not others.
Identity is a multi-faceted thing. We humans tend to construct our own view of who we are and pick those aspects of ourselves which we regard as most telling. These identities might be any mixture of sex, place of birth, job, friends, philosophical or political beliefs or character traits. Parents and sibling usually get a look in too. Many readers of this magazine will probably be of the opinion that their family tree – their genealogical ancestry – is not only fascinating in itself but can also provide meaningful information about ‘who we are’. Some will want to go further and delve, as far as science and pockets will allow, into their genetic ancestry. But what is the relationship between genealogical and genetic inheritance?
As I explained in a recent article in Family Tree, once you are conceived genealogical ancestry is a completely deterministic thing. In genealogical terms you are without any doubt descended from or related to your ancestors in a definite way. I explained why the number of your direct ancestors (parents, grandparents etc) doesn’t simply double in each generation: it’s because of inbreeding and the resultant ‘Pedigree Collapse’. But if we put this to one side here, you are descended one half from each of your parents and one quarter from each of your grandparents and so on. If you could accurately identify all your ancestors you could calculate the precise mathematical genealogical relationship between you and any one of them. One measure of relationship is called the Coefficient of Relationship. This would be 50 per cent between parents and children, 25 percent between half siblings and only 3.13 per cent between second cousins. However this measure can be unrealistic because it assumes zero relatedness on other lineages, which, as I discussed in my previous article, is not the case.
In terms of identity, if you had four Scottish great grandparents, two Russian great grandparents, one French great grandparent and one Japanese great grandparent, then you could perfectly validly say you were genealogically, and maybe culturally and linguistically too, one half Scottish, one quarter Russian, one eight French and one eighth Japanese. But does the same hold true for your genetic inheritance? The answer is ‘not quite’. To understand why we need to understand a little about human reproduction and how genes are passed from generation to generation.
Genes and reproduction
Humans have 23 pairs of chromosomes, making 46 in total. These contain all our genetic information. Two chromosomes determine sex – you get and X or a Y from your father and an X from your mother. That leaves 22 other pairs of non-sex (‘autosomal’) homologous chromosomes. Homologous simply means that while each half of the pair has the same length, basically the same functions and indeed the same genes, the pairs of genes can appear in different versions – called alleles. A well known example of this is found on chromosome 15, where one gene (allele) can either code for the expression of brown or blue eyes. (Note: non-sex chromosomes are simply numbered from 1 to 22: 1 being the longest, 2 the second longest and so on.) Having 46 chromosomes (or 23 pairs of homologous chromosomes if you prefer) is one of the defining characteristics of being human. Chimps have 48 and dogs 78. If by chance you get more or less than 46, severe health problems can arise. An extra number 21 chromosome for example, i.e. a triple rather than a pair, gives 47 chromosomes and results in Down’s syndrome.
I hope it’s clear that if each parent has 46 chromosomes any child must also have 46. Thus during the process of reproduction the combined number must be halved – and indeed it is.
Let us consider any one of the 22 non-sex chromosomes, for example number 15, which as I mentioned codes for eye colour among other things. See the image which represents the pairs of ‘number 15’ homologous chromosomes for one individual and his/her parents and his/her grandparents. I’ve given each part of the chromosome pairs a different colour and just for illustrative purposes assume that they are passed down unchanged (which they aren’t). In this example the individual is red & blue. He/she has inherited the red part of his/her paired chromosome 15 from the father and the blue part from the mother: 50 per cent from each of the parents as we might expect and with the required reduction. The father has, here, the red plus green combination and there was an independent 50/50 chance of the child getting either red or green from him. The same applies to the mother with blue and yellow. Thus the red & blue combination is only one out of four possible combinations which could be inherited from the parents. And so it is with all the other 21 non-sex chromosomes, although graphically we’d want different colours for each to differentiate them all. Thus in total we’d get 50 per cent of our total genetic inheritance from each parent.
But consider just the paternal line for a moment. You can see that the father could equally as easily have inherited any one of four different colour combinations from his parents: green & red, green & orange, pink & red and pink & orange. There are also four combinations on the maternal side. This means that given the number 15 chromosome combinations the grandparents had there was only a 1/16th chance of this individual having got the red & blue combination – 1/4×1/4 – and a 15/16ths chance of any other combination. It might also be of interest to note that taking all the chromosomes into account there are over 8 million possible combinations of chromosomes (2 to the power 23) from either your father or your mother!
If humans reproduced in this way (they don’t) you can see that you would have inherited genes on chromosome 15 from only one of your two paternal grandparents and only one of your two maternal grandparents, and none whatsoever from the others. Perhaps surprisingly you would also have inherited genes on this chromosome, once again, from only two of your eight great grandparents. In fact you would have chromosome 15 genes from only 2 ancestors in any generation. Of course, because there are 22 non-sex chromosomes, the particular pair of ancestors you might have inherited genes from, on each chromosome in each generation, will likely be different. An interesting thought is that if humans reproduced like this we would all have a maximum of 46 distinct genetic ancestors however far you go back (2×23). The vast bulk of your genealogical ancestors wouldn’t be genetic ancestors at all!
Shuffling the pack
Luckily for biological diversity, natural selection and human health, something else happens when we reproduce. Not only are chromosomes independently assorted and their number reduced by half, as in the hypothetical example above, but, in addition, before your mother and father each pass on half of a chromosome pair to their sex-cells – called gametes: eggs in females and sperm in males – some genes on each chromosome are shuffled. Individual genes (alleles) on ‘opposite sides’ of the chromosome cross-over or recombine. This occurs when sex cells are being formed in a complicated multi-stage process. The homologous chromosome pairs first double and then, in a two-step process known as meiosis, chromosomes join, some genes then ‘cross over’ or ‘recombine’, then the chromosomes segregate again. See the second illustration. In males we end up with four separate sperm cells each containing 22 different ‘haploid daughter chromatids’ – this just means one half of a pair – plus the sex chromosome. For females it’s a little different. They end up with just one fertilizable egg, again containing 22 haploid daughter chromatids plus the sex chromosome. Three other potential eggs, called polar bodies, become redundant. One sperm will fertilise one egg to create a new person and we’re back to 46 chromosomes again, but very different ones.
How likely two genes are to cross-over is a probabilistic process and depends in large part on how far apart they are on the chromosome; the nearer they are (the more ‘linked’) the lower the probability of crossing over. Actually in humans the amount of gene shuffling is minimal, quite often being as low as one gene cross-over per chromosome; other times only two or three. Even with such genetic shuffling, it still means that any individual will still get exactly 50 per cent of their genes from each of their parents (both on each chromosome and in total), but they need not, and probably will not, inherit 25 per cent of their genes from each of their four grandparents – again on each chromosome or in total. While our best guess will be 25 per cent, 25 per cent, 25 per cent, 25 per cent, like all averages based on probability there is a wide range of possible results. Imagine tossing a coin four times. Before you start the best guess would be that you will get two heads and two tails. But you could also quite conceivably get three or even four heads. If you have a few goes it won’t be too long before you actually witness this. What is more, if after three tosses you have got three heads, while the probability of getting a fourth head is still 50 per cent – because it’s independent of anything that went before – having got three heads first, after the fourth toss the only two possible final results are 3 heads and a tail or four heads! The cumulative outcome is dependent on what went before – as it is in genetics.
What’s the answer?
To put the outcome in a nutshell: while in any large population the average percentage of genes inherited from each and every grandparental generation will likely be very close indeed to 25 per cent (or 12.5 per cent for great grandparents), for any single individual the probability of them having exactly 25 per cent from each of their own four grandparents is far less than them not having 25 per cent – i.e. having any other proportion at all that is more or less than 25 per cent. On any particular chromosome, which might contain genetic ‘codes’ for particular physical or behavioural traits, I hope you can see that it is quite possible, even quite frequent, that you have inherited very, very little genetic information, maybe even none, from a grandparent or great grandparent. On the other hand it’s highly unlikely, though still remotely possible, that in total you will get almost or exactly no genes from any one of these relatively recent ancestors. But as you go further back in your ancestry the likelihood of having inherited no genes from a remote genealogical ancestor becomes more significant.
Finding your genetic ancestry
Moving away from theory and towards what we find in the real world. Some companies now offer genetic inheritance tests. There is a whole new industry called ‘Genetic Genealogy’. Most well known are tests using mitochondrial DNA. This is DNA situated outside the nucleus of a woman’s egg and is passed unchanged from mother to daughter except for random mutations. Males also get mitochondrial DNA but can’t pass it on. Another popular test follows the male Y chromosome, passed more or less unchanged, except for mutations, from father to son. The results of such tests are interesting but they only tell us something about two single genetic lines out of our hundreds of such lines: those of our mother’s mother’s mother etc and our father’s father’s father etc. More recently tests of our non-sex genetic inheritance have become available. These are more complicated than with the Y chromosome and mitochondrial DNA because these genes are constantly being shuffled. As genetic science progresses such ‘autosomal’ DNA tests are becoming more and more informative. Remember, with the exception of some non-sex inheritance on the X chromosome (like colour blindness), everything else, according to conventional biology and genetics – ignoring ‘epigenetics’ – comes from these non-sex chromosomes: physical, mental and behavioural characteristics for example.
There are various studies of such autosomal genetic tests and, although the numbers differ, they all clearly show that there is a significant range in terms of genetic inheritance. One example being what percentage of our genes we get from each of our grandparents or great grandparents. The highest percentage of genes received by a person from a grandparent that I’ve so far seen reported is 31.5 per cent, which of course means the other grandparent contributed only 18.5 per cent.
Genetic and genealogical ancestries are not the same. You or I will most likely have at least some genes from most of our ancestors, but how much will vary quite a lot, as will which mix of genes and traits we inherited. Returning to the example of Scottish, Russian, French and Japanese ancestry I started with. It is in fact highly unlikely that the genetic ancestry ratios will match the genealogical ones. Some of the proportions or percentages could be significantly higher and some much lower – as long as they add up to 100% of course. You might genealogically be one eighth Japanese but genetically you’ll most probably not be. And, what is more, whether you did or didn’t get any particular genetically carried trait, or even talent, from your Japanese ancestor is basically just pot luck.