What are the odds that two opposite-sex siblings are genetically unrelated (assuming we exclude mitochondrial DNA)?
As I understand genetics, it is possible in principle for a brother and sister to receive completely different chromosomes from the same parents (i.e., the brother gets all the Mom’s “left-side” chromosomes and the sister gets all the “right-side” and same with Dad’s chromosomes). This isn’t possible for same-sex siblings, since they either share the same X chromosome from Dad or the same Y from Dad. And all mitochondrial DNA comes from Mom.
But what are the odds? Is it just .5 to the 22nd power? Or is there biology I’m missing here?
Edit: to clarify, obviously they share a ton of genes because they are both human, but I’m talking about the odds of not having gotten the gene from a common parent
Assuming I’m not misunderstanding the question I think what you are missing is that X and Y chromosomes are only one of twenty three contributed by each parent so they would still share genetics.
I’m going to to say the odds are essentially zero. Because homologous recombination happens in lumps, not at the individual gene level. So, you don’t get just a random shuffling of all the genes.
A more likely scenario (although we have not way of calculating these odds) is that recombination never takes place at all, and the chromosomes are passed on intact. And you’d have to have this happen twice to the same parents such that different chromosome pairs are passed to the offspring.
Homologous recombination is an ancient evolutionary adaptation. I would suspect, although I’m not completely sure, that a mutation that inhibited that process would have all sorts of adverse effects such that the organism wouldn’t survive.
ETA: I’m sure that’s not the answer you were hoping for.
It is quite likely that siblings are genetically unrelated. My kids are only related as cousins, but it was entirely possible that they would be biologically unrelated. I suppose this thread is about siblings with the same biological parents though.
I assume by “crossing over” (cj) and " homologous recombination" (jm) you’re both talking about the recombination of each chromosome that occurs among the parent’s chromosomes in meiosis. If so, it isn’t clear how that it relevant to my question, which is about the likelihood of siblings sharing the same gene passed from a parent.
Can you spell it out for me how this would mean the two siblings would necessarily share any genes from a parent?
It may be that my question wasn’t clear enough. Here’s my assumptions and what I’m asking:
For any given chromosome pair (setting aside the sex chromosomes for the moment), let’s say Dad has one chromosome that’s ABCD and its pair which is 1234. Recombination occurs during meiosis, and we get two sperm, perhaps A23D and 1BC4. Similarly, Mom has EFGH and 5678, with two eggs that are EF78 and 56GH.
Now, we could have two siblings, one gets A23D and EF78, and the other gets 1BC4 and 56GH. Thus, none of their genes for this chromosome are overlapping. Right?
Also, each parent is going to have many homozygous alleles, so all the spermatozoa or ova are going to have many genes identical to all the other spermatozoa or ova.
Well, yes, obviously. Both are human. Hence my clarification: obviously they share a ton of genes because they are both human, but I’m talking about the odds of not having gotten the gene from a common parent.
As most people have already said, the possibility of having non-related siblings is so low as to essentially be zero because of homologous recombination.
I would like to point out however that there are some organisms which do not exhibit homologous recombination. So in their case it would be possible.
Can you explain this further to me, because I’m just not getting it. Why should shuffling within one parent’s process of meiosis affect the odds of sibling relationship?
Oh, I see why. Because I was thinking of the question in terms of sperm and eggs from one round of meiosis. But in order to have no overlap from different rounds, the recombination would have to be nearly identical. Got it.
But so the odds are merely low, not impossible, right?
So, lets say we lined up all 23 pairs of chromosomes for each parent. That would be 46 total.
Now give the ‘right’ pair of each parent to Daughter. And the 'left" chromosome pair to Son.
they would be genetically unrelated right?
The problem is with homologous recombination, this scenario would never happen. Because the left and right chromosome swap bits. You’ll always pass along chromosomes which are part right pair and part left pair. To end up with unrelated siblings in this scenario, the the two sibling would have to have exact opposite series of homologous recombination events.
This just wouldn’t ever happen.
Considering .5^22 is already equal to 0.0000002384185791015625. That’s 1 in ~5 million.
I’m sure having the exact opposite homologous recombination events would be exponentially smaller. You’d need to know recombination frequencies and magically restrict them to specific portions of the chromosomes in order to even have a chance at calculating it.
My wild ass guess? It’s less then in 1 in 10 billion. So it probably has never happened.
I’m sure the probability would be infinitesimal even without homologous recombination. I don’t think anyone is denying that the odds are astronomically low, but like the OP, I’m still interested in knowing if this very low probability can be quantified, even if only by an upper bound.
As I said earlier, the chromosomes don’t exchange individual genes randomly. They exchange stretches of genes, so it’s not like every gene has an equal chance of crossing over, unrelated to whether the other genes cross over, too. Your simplified model is not representative of what happens in the real world, where each chromosome has long stretches of genes on it.
But so the chance is non-zero, right? Do** John Mace** and california jobcase agree?
Another version of the question would be: what are the odds that biological siblings are no more related than biological cousins? Do both question turn on quantifying homologous recombination odds?
Indeed, the fact that recombinations involves chunks of genes raises the statistical likelihood of unrelated siblings, all else being equal.
Having the same genes at the same loci on homologous chromosomes is usually considered genetically related. Your idea of unrelated siblings may not have an absolute zero chance in theory, but in practice I’d bet on a zero chance every time 'til infinity.
An almost identical question was asked and partially answered in an earlier thread.
I found the above Googling to recover an earlier-posted fact:
“… during meiosis chromosome pairs will crossover (Chromosomal crossover - Wikipedia) typically 2 to 4 times per pair.”
I don’t know the statistics on crossover-point selection, but, ignoring the non-recombining scenario posited by John Mace, Zero would seem to be a very excellent approximation to the answer to OP’s question.
Meh. I wouldn’t call it my idea. I think the folk notion of “relatedness” in the context of families is not just happen to share some common genes but instead share some common human genes as a consequence of having inherited them from a family line.
And in the context of human births, it isn’t clear to me why odds in the range of 1 in 10 billion are functionally the same as zero. Those odds would suggest there have been at least one set of human siblings who were no more closely genetically related than their parents were to each other, wouldn’t they?
