I don’t have the time now to address this at length, but I think where we’re getting tangled up is in the difference between genetically similar and related. You could, theoretically, have two people who are genetically identical yet completely unrelated through random chance. My understanding is that the OP was asking about relatedness. My interpretation of that was looking at where the genetic information came from, rather than what it was. In other words, for this question, we’re treating a paternal chromosome as a completely different thing from a maternal chromosome, although of course we know they aren’t.
I don’t think this question is answerable without specific sequencing of each genome, but I know the answer would rely in SNPs, or single nucleotide polymorphisms. A SNP would be the mother having an A on one chromosome, maybe an A on the allele, and the father having Ts at the same spot on his two alleles. Of course, the father could be A/T or A/A as well.
There are ~3 million SNPs in the genome out of 3 billion base pairs, which means at maximum 0.1% of the genome is different between individuals. Granted, there are other variations like VNTRs, but they should work in the same manner, and SNPs account for 90% of the differences. So you need to specifically look at each person at each SNP to determine the relationship. Also, I do not believe there is a single SNP (or at least one has not been found) that has more than 2 nucleotides at a position.
For example, lets look at a single SNP and how it relates to the son and sibling:
There are 9 possible combinations of mother/father at a single nucleotide (considering 2 alleles):
Number 1…2…3…4…5…6…7…8…9…
Mother AA AA AA AT AT AT TT TT TT
Father TT AT AA TT AT AA TT AT AA
Son/sibling possibilities for each scenario:
- AT (100% chance brothers are identical)
- AA or AT (50%)
- AA (100%)
- AT or TT (50%)
- AA, AT, or TT (33%)
- AA or AT (50%)
- TT (100%)
- AT or TT (50%)
- AT (100%)
- 0%
- 50%
- 100%
- 50%
- 33%
- 50%
- 100%
- 50%
- 0%
So the only difference is at scenario #1 and #9 where the father/son are 0% identical for the SNP. But what confuses me is the “identical” vs. “similar”. The father still has 1 copy of the nucleotide of the son in both of those scenarios, so they are still 50% similar, correct (even though it’s a 0% identical)? Brothers may have a 50% chance to have the same SNP AA or AT, but still brother #1 has 2 copies of A, and brother #2 has 1 copy of A and 1 of T…does this make them 50% identical, or 75% identical?
I really need someone with a probability background to analyze my data.
Overall, considering most people tend to marry people with similar ethnic backgrounds (and therefore similar SNPs), I will guess that brothers will be overall slightly more similar than fathers/sons.
Ok, Bob55 I think you have the right idea. You made one error though, AT x AT gives (AA,AT,AT,TT), so the chance of father-son similarity is 50% not 33%. Also as you note, we should look at the similarity, not at the chance that they are identical.
I will give a difference of one nucleotide the value 1. If there is 50% there is a 1 nucleotide difference, that will get the value 0.5. A difference of 2 nucleotides will get the value x.
Case - father/son - brother/brother
1…1…0
2…½…½
3…0…0
4…½…½
5…½…½ + 1/8x
6…½…½
7…0…0
8…½…½
9…1…0
So in order to look at the difference, we can focus on case 1,5 and 9.
Now I assume that the population is in Hardy-Weinberg equilibrium.
Thus:
A has the frequence a
T has the frequence t
AA has the frequency a^2
AT has the frequency 2at
TT has the frequency t^2
Case 1 will happen if AA meets TT, which happens with w frequency of a^2*t^2
Case 5, AT meets AT: 2at * 2at = 4a^2t^2
case 9, TT meets AA: a^2t^2
So, to total up. the extra difference for father son is:
(case 1 + case 9) * 1 = 2a^2t^2
The extra difference for brother/brother is:
(case 5) * 1/8x = 4a^2t^21/8x =½a^2t^2x
So it all boils down too: How different are two people who have both nucleotides different? I personally have no idea. If we assume that they are twice as different as two people with one nucleotide different, then x = 2, and we have:
father/son diff:
2a^2t^2
brother/brother diff:
½a^2t^22 = a^2t^2
So in that case, fathers and sons are more different overall, by a small degree.
(I hope this is possible to understand. I’m too lazy to explain thoroughly.)
And I disagree.
He said “genetically closer” in both the title and content of the OP.
Say a mother and father with brown eyes produce a child with blue eyes, because they were both carrying a recessive blue eye gene. It seems to me utterly misleading and meaningless to say this gene is 50% the same as the father’s and 50% the same as the mother’s.
More importantly though, this interpretation of relatedness seems to imply that any two siblings are 0% related, since neither sibling obtained any genetic material from their sibling.
To interpret sibling similarity in any other way would be inconsistent.
Hey, this isn’t nearly complicated enough yet. Let me add a couple of things.
First, you need to account for the fact that the X chromosome is larger than the Y, so the boys are actually more related to their mother than their father. Girls will, at this level, be equally related to both.
But hold on, you need to also account for mitochondrial DNA, of which there is a single copy, inherited from the mother. So even the girls are slightly more related to their mother than their father. In a theoretical sense anyway. There isn’t a whole lot of variation in m-DNA, and I don’t think it’s ever been shown to create any real difference in organisms.
[Finally, there’s a possibility that non-DNA material inherited in the egg may have an effect on the child. I don’t know if this has been found anywhere yet. It would mean slightly more relatedness to the mother again]
Those are valid points. The first one doesn’t matter in the father/son vs brother/brother comparison though. But the second one does. Brothers will have the same mitochondria while the father will never have the same as his sons. (Unless they happen to be identical while not inherited of course.)