In the process of working out a family tree on a well-known website dedicated to that purpose, I met a second cousin I had not known about (although he had an inkling of my existence). His grandmother and my grandfather were brother and sister; thus, on our family tree we have two great-grandparents in common (and six each that are different). So one-fourth of our “pedigree” is the same. Yet Wikipedia says second cousins only have a 3.13% relatedness (non-identical siblings, by contrast, having a 50% relatedness), and other sites I’ve checked say the same.
I wasn’t expecting the relatedness to be 25%, mind you; given that we only have one-fourth our ancestry in common as compared to the 100% two siblings have in common, I was expecting it to be more like 12.5%. But I guess I just don’t see how we can have such a significant share of the family tree in common, and yet be so distant in relatedness. By the same token, it strikes me as bizarre that when my kids and my sister’s kids have their own children, my sister and I, who grew up together with the same biological parents, will have grandchildren who are so distantly related.
Roughly speaking, with each step (sibling-sibling or parent-child), only half the genes are shared. From you to your second cousin is 5 steps: your father’s father’s sister’s daughter’s son (or whatever the sexes are). 0.50.50.50.50.5=0.03125, or 3.12%.
With respect to your great-grandfather in common, you inherit one eighth from that great grandfather. Your cousin also inherits 1/8th from that same great grandfather. But you didn’t both inherit the same eighth. So what you have in common from great-grandpa is one eighth times one eighth, or somewhere between 1.5 and 1.6%.
The same can be said of great-grandma, so you add those 1.55% factors and get roughly 3.1%, .
But then where do all the other genes come from? I mean, each of us (humans) has at most eight great-grandparents (one can have fewer if cousins marry etc.). And all of our genes have to come from (or if you prefer, through) them–there’s nowhere else to look. So if he and I have two great-grandparents in common, but only just over 3 percent of genes in common (not including those everyone has in common of course, or those that might have bounced back around from more distant relations we’re not aware of), then…what, I’m getting the other 97 percent from the other six grandparents?!? That still makes no sense to me.
I need an explanation that goes further than multiplying decimals, I’m afraid. (I’m more than capable of comprehending arithmetic, having once been a math major–doing well in two semesters of calc and discrete math before dropping out of the program because I couldn’t comprehend the non-numerical “foundations of geometry” proofs.)
The percentage of relatedness is not the same as sharing that percentage of genes.
Part of the issue comes in with crossover during meiosis. Not all the genes you inherited from your great grand parents are the same as the ones your cousin inherited, even if it’s the same gene. You and your cousin could have got different alleles. Even if you got the same allele, the aforementioned mentioned crossover is problematic. Combine this with the multiple generations between you, and how close you are related is going to get pretty slim.
The genes themselves are the same among all people (essentially).
However, the versions of the genes you inherit vary, and the specific combinations of those versions make each of us unique, genetically. So all of us have genes for blood type and skin color and metabolic enzymes etc. But the version I have of a particular gene may differ from the version you have. And in the 30,000 genes humans have the specific combinations are more similar among related people.
A brother and sister from the same parents will only have 50% of the genes in common. (on average, give or take, it’s random).
The cousins, offspring of those, have 50% from their respective related parent. So one cousin has 50% of their genes from grandparents A, as will the other grandchild-cousin. But, these on average are 50% matching, 50% non-matching genes of the 50% of parent genes that match; so the cousins will have 12.5% matching genes.
Another generation, the generation that have 2 great-grandparents in common - 3.13%.
You get half the genes from your one parent, but only half those are likely to be matching genes. So each generation diverges to 1/4 of possibly matching genes. Only the first generation - 2 common parents - gets half matching, a quarter matching from each parent.
YoDoc, IvoryTower, Sailboat, John Mace: I guess that the rest of us are playing a game where we ignore crossover, common elements of the human genome, and sex-linked chromosomes here. We talk about it in terms of genetic inheritance, but it’s still a huge simplification, and gets complicated by the intrusions of actual genetic science:
Rules of the game, as I understand them:
All children inherit 50% of genes from each parent
All inheritances are independent
If genes weren’t inherited from a common ancestor, they’re assumed to be ‘not the same.’
I would have put it this way - my brother is closer genetically to a chimp than to me.
Basic genetic material - what determines how many legs, kidneys, brain cells and noses we have - that is in some way common to all animals, and in some ay divergent between species. So we all have blood genes; human genes are somewhat different from chimps. Human Blood Type A and Type O genes are almost the same, but slightly different, and thereare subtle differences between one gene and another that do the same thing in humans.
IIRC (IANAbiologist) Every time cells split, the DNA strands are duplicated; sometimes this introduces errors. If the errors are to bad, the cell does nt work, or you get cancer, or whatever. If the changes are harmless, they pile up. So all human DNA is different, and the longer between common ancestor, the more likely that section of the DNA is more different. This is how they say “this population split from this X thousand years ago.”
There is “garbage DNA” filler between the sections (genes) that actually do things - IIRC over 90% of DNA is filler. The number of repeating sections of this filler DNA can change quite a bit without affecting function. DNA tests chop at a certain place, then measure how long the filler DNA is. Different DNA has different numbers; these can change, but rarely enough, odds are your “number” at a gene site matches your ancestors several generations back. Given distributions, the odds of a paricular site being, let’s say, a 7 or a 3 is, let’s say, 1 in 10; pick 2 such sites, if they match a parent, odds would be, say,1 in 100 you go it from the same parent. By the time you pick 16 such sites, the odds that you do not match one parent or the other are astronomical. The actual odds by site are determined by gathering statistics for the general population.
You are confusing types of genes with versions of those genes. Chimps v humans share about 96% of the types of genes. There are enough significant gene differences to consider those 4% completely different genes.
My kids and I (and basically every human) share 100% of types of genes. However we vary from each other on the versions (blond hair v black hair) of those genes.
So, my kids and I share many many of the same versions of genes (we’re all green-eyed, for example), more so than I do with a stranger (who may be brown eyed).
The terminology is confusing. When we say two second cousins have a 3.125% degree of relatedness, it means that is the chance that any particular one of their genes was inherited from the same ancestor.
Or, you can think of it as the probability that they share a rare (non-sexed-linked) gene. So if their shared great-grandfather had a unique gene, there’s a 1/64 chance that both second cousins have that gene. And likewise for their shared great-grandmother.
This is important for determining the risks of inbreeding–the chance of a child inheriting two copies of a gene from the same ancestor is 1/4 the degree of relatedness of their parents. So, for the child whose parents are second cousins has 0.78125% chance of having two copies of a gene from one ancestor. That’s a problem if the gene is recessive and disease-causing. (For first cousins, that’s a 3.125% chance, and 12.5% for siblings. Thus the widespread prevalence of taboos against sibling parents, some prevalence against first-cousin parents, and little prevalence against second-cousin parents.)
Comparing the absolute genetic similarity between individuals (or species) is counting the fraction of genes that are actually the same. There’s no direct accounting for ancestry.
I think the point here that some people (not necessarily anyone in this thread) don’t understand is how genes get passed down to sex cells (eggs and sperm).
A normal human cell has 23 pairs of chromosomes, for a total of 46. The sex cells get half of each set, for only 23 total. But which half of each pair is randomly selected. So, working from one genome, there are 2[sup]23[/sup] possible sex cell genomes. Two different children are never going to get identical sets, statistically.
That’s why you and your cousin don’t end up with the same inheritance from the same great grandfather.
There’s another process - from what I understand, when the genes duplicate during cell division, the two chromosomes may exchange segments - as a result, you do not pass on the same complete chromosome to your descendants that you got from your parents; it may consist of chunks from both parents.
The only exception, that is easy to track, is the Y chromosome that detemines “male”. (and so presumably, the corresponding X in men) It does not have a match to swap with, so it only changes very slowly due to occasional mutation. A direct male descendant will have the same Y as his father, grandfather, great-grandfather, etc. Small differences creep in over time, so how exact the match is can tell you how far back their common anecstor is.
Similarly, there are small rings of DNA in cells that are not part of the nucleus/chromosomes. These mitochrodrial DNA are only passed on in the egg cell, so are only inherited from the mother. DNA tests on these determine people who share a common maternal line; and again, how different this DNA is between two samples will tell you how long ago two populations diverged.
It doesn’t look like anyone has really explained the math behind it, so I’ll give it a go. You have two great-grandparents in common with this second cousin which accounts for 1/4 of your inherited genes. I’m going to assume the other 3/4 isn’t related, even though technically if you go back far enough it is. From each of your great-grandparents, you inherited 1/8 of your genes. However, you didn’t inherit the same genes. For each of the genes you inherited from this one great-grandparent, there’s a 1/8 chance that your second cousin has inherited this same gene. Therefore, on average, you will share 1/64 of the genes from this great-grandparent. Then you double that since you have two great-grandparents in common and you get 1/32, which is roughly 3%.
Chrisk, it looks like you posted while I was composing my last post. Yes, your explanation did make sense, thank you.
So when I see aspects of my second cousin’s face that remind me of my dad, or when my sister sees a photo of his mom (my dad’s first cousin so our first cousin once removed) and says she looks like me “in drag”, are we just kidding ourselves? Or is 6-odd percent enough for strong resemblances in some cases? (I suppose it’s also possible that we share more relatedness than that too since it’s just an average.)