I probably need a primer in genetics...

[jmcvann]

I was just reading this blog post - addmittedly old, but interesting.

http://lesswrong.com/lw/kr/an_alien_god/

But when I got to this sentence: "Your brother or sister shares half your genes," I paused. Is that correct?

Is this one of those things I only think I know? (There's probably lots of those things!) You get half your genes from mom and half from dad? But the express differently, right? So my sister may have brown eyes while I have blue eyes.

As I continued pondering this I came to wonder, is it possible that my hypothetical brother could get exactly the "other" half of dad's genes and the "other" half of mom's? Could we appear to not be genetically related at all?

I confess up front that genetics is not my forte. I'm posting here to learn. Teach me! Please!! (Reading suggestions welcome)

 

[Professor Yaffle]

The sentence really should include the words "on average".

 

[paximperium]

I'm a bit rusty but here a simple rundown. The genetic "sharing" when reproducing is via Chromosomes and a multitude of complex encodings of the genes.

You have 23 pairs of chromosomes(46 total) with 23 of your dad's genes(including Y) and 23 of your mom's Chromosomes (including X). However, the pairs of Chromosomes from your dad are not shared as per a mirror image from your siblings.

For example you have a 50% chance of getting Chromosome 1a or 1b while your brother of sister has a 50% chance of getting either 1a or 1b either. The same with 2a and 2b. For you could have 1a2a while you sister could have 1b2b or 1a2b etc.

So do the math. The chances of you sharing the same sets of chromosomes between yourself and your siblings is definitely not fifty-fifty although it would average out in that way, in fact there is a small chance you could share no genetic similarities between yourself and your sister(remember you and your brother always share a Y chromosome) at all since your sister could receive Chromosomes from both mom and dad that you did not get. About 2*22...very small chance.

You share 50% of your mom and your dad's Chromosome(with the exception for some minor encoding differences) but amount of genetic similarities between yourself and your siblings are based on chance but on average should be about 50%(unless you count in the sex chromosomes)

Anyone else with more recent genetic knowledge please chime in.

 

[jmcvann]

Other questions: Would the dominant/recessive value of any particular gene make the genetically opposite sibling impossible? And does dominant always win? If dad "wants" my eyes to be brown and mom "wants" my eyes to be green, will they be brown for sure? (I'm pretending to know that brown is dominant over green!) Or does the recessive gene win a battle now and again?

ETA: Read "gene" as "chromosome" (right?)

 

[sanguine]

The sentence really should include the words "on average".

This is the simplest and best fix to that sentence.

If anyone cares about the detalia:

A brother and sister, by definition, won't have all the same genes because there are some genes that only appear on the Y chromosome, which is inherited from the father, and solely present (assuming no assortment defects) in the brother in our theoretical brother-sister pair.

Otherwise, in terms of nuclear DNA -- that is, the DNA the your other 22 autosomal chromosomes consist of -- then you have a very, very small chance of getting exactly the opposite 22 chromosomes as your sibling, with no internal crossover events.

Note that last sentence -- chromosomes from parents don't just cleanly keep getting passed down. There are crossover events where the equivalent parts of the matching chromosomes (say, part of the short arm of chromosome 8) swap. Thus, each individual chromosome that you and your theoretical sibling are comparing is most likely some blend of the parent chromosomes -- for example, you may have a chromosome 8 that is 90% mom and 10% dad, and your sister may have a chromosome 8 that's 50% mom, 50% dad.

All of these means that the odds of two children from the same parents having no matching nuclear DNA is vanishingly small.

Incidentally, I keep saying "nuclear DNA" because you and all your siblings, regardless of gender, received your mitochondria from your mom. As a consequence, your mitochondrial DNA is derived from hers, and you and your siblings are likely to be very similar (or even exactly the same) in terms of mitochondrial DNA.

 

[Ziggurat]

Is this one of those things I only think I know? (There's probably lots of those things!) You get half your genes from mom and half from dad?

Yes, but the bit you're missing is that you only get half of your dad's genes. Which half is basically random: each sperm has half the dad's genome, and any two sperm will, on average, share half of that half in common. Since the number of genes is very large, the deviation from average is pretty small. Same thing goes for the eggs and the mom.

So for example, let's consider you and one sibling. 25% of your genes come from mom but are not shared with that sibling. 25% of your genes come from mom and are shared with that sibling. 25% of your genes come from dad and are not shared with that sibling. 25% of your genes come from dad and are shared with that sibling. Same percentage breakdowns with other siblings, but it won't be the same genes making up the percentages (ie, you could share a gene with one sibling but not another).

For a given gene from a parent with n children, the chance that at least 1 offspring carries that gene is 1-(1/2)ⁿ. The chance that all offspring carry that gene is (1/2)ⁿ.

Of course, all of this is ignoring the X and Y chromosome issues, which don't act the same way since there's no interchange of genes between X's and Y's (but there is between two X's).

 

[paximperium]

ETA: Read "gene" as "chromosome" (right?)

Genes and Chromosomes are different things.
Genes are somewhat colloquial term. Basically a gene is the most basic inherited unit so it can mean a single protein or even amino acid or be more complicated.

Chromosome are the large structures that DNA conglomerate under that genes are clustered under(eg. Gene A in located in Chromosome 4). They are visible under a microscope. They are the basic structures involved in cell replication and reproduction.

 

[sanguine]

Other questions: Would the dominant/recessive value of any particular gene make the genetically opposite sibling impossible? And does dominant always win? If dad "wants" my eyes to be brown and mom "wants" my eyes to be green, will they be brown for sure? (I'm pretending to know that brown is dominant over green!) Or does the recessive gene win a battle now and again?

ETA: Read "gene" as "chromosome" (right?)

No, read "gene" as "gene" here.

"Dominant" and "recessive" are kind of blunt instrument terms that come from early genetics, where we only had a readout on the final outcome (phenotype) of genes.

At the level of the actual gene, there are a number of reasons why a gene can yield a "dominant" or "recessive" trait, and there can be degrees of dominance. The simplest case, for example, might be comparing alleles (different gene variants) of a pigmentation gene. We could imagine that the "green" eye color actually results from the "brown" gene being nonfunctional, letting other pigments be visible. Then the "brown" allele will always win because it's a case of the presence of absence of brown pigment.

Note that I'm making that example up; I largely work in metabolism and don't know how eye color works.

However, it's possible for different alleles to interact in more subtle ways. What if, for example, eye color depended on different gene regulators as well? We could imagine that "brown" pigment gene beats "green" pigment gene, but there could also be a gene regulator that simply turns off the brown gene when it's present. Thus, inheriting:

BROWN and ON

along with

GREEN and OFF

Would yield green eyes again.

To return to your original question, dominance has no bearing on the actual genes. Again, it's a readout of what those genes do, not their presence or absence. The only thing that's required for one off the offspring to have the "recessive" gene is for them to inherit it.

It is possible that the collection of alleles for a given gene present in the parents will make a certain phenotype impossible. If we return to the simple initial case of green versus brown that I gave above, if mom is GREEN-BROWN and dad is BROWN-BROWN, then barring mutation, none of the kids can have green eyes -- but our theoretical siblings can still inherit entirely distinct alleles from each other (you just won't be able to tell by looking at them).

 

[Ziggurat]

Other questions: Would the dominant/recessive value of any particular gene make the genetically opposite sibling impossible? And does dominant always win?

Dominant always wins, which is why it's called dominant. For truly recessive genes, expression will only result when both chromosomes carry that same gene. So both parents must have at least one copy of the recessive gene (again, ignoring sex-linked genes, since males have only one X chromosome) in order for offspring to express the gene. But this also means that offspring can express a gene that neither parent does, if the parents both have one recessive and one dominant gene.

But for many genes, there is no simple dominant/recessive duality, expression can get much more subtle than that. And many traits (for example, height) do not depend on single genes. So in general, these guidelines aren't really enough to tell you a lot about gene expression.

ETA: Read "gene" as "chromosome" (right?)

No. Chromosomes contain many, many genes. They are scrambled when being passed to offspring. You definitely mean genes, not chromosomes

 

[paximperium]

Other questions: Would the dominant/recessive value of any particular gene make the genetically opposite sibling impossible? And does dominant always win? If dad "wants" my eyes to be brown and mom "wants" my eyes to be green, will they be brown for sure? (I'm pretending to know that brown is dominant over green!) Or does the recessive gene win a battle now and again?

Dominance is complicated. It is often just the expression of the Gene(Genotype) to the Phenotype(physical characteristics).

For instance: Say the gene for Protein A(A) is dominant while Protein b(b) is recessive. Protein A causes normal clotting while Protein b is abnormal clotting. If you get AA, you clot as normal. If you get Ab, you clot as normal. If you get bb, you have a Clotting disease. So Protein A seems to win.

However, if you measure Protein A in Ab people, you will notice that it is 10% less than AA people and some variations will be 50% less. This is the case with people with Sickle Cell trait/disease for instance. In people with SS, they have the disease, in Ss, they have abnormal cells but no disease. This leads to a huge variety ranging from co-dominance, incomplete Dominence, etc.

Conclusion: It's complicated. It depends on the trait.

 

[sanguine]

Just for clarity's sake, since the dominant versus recessive thing can be confusing:

It's helpful to keep in mind that when we discuss "dominant" and "recessive" genes, we're referring to different alleles (variants) of the same gene, and the fact that the "dominant" one will yield a visible physiological effect - a phenotype - that trumps the one the "recessive" one yields.

Thus, if we call the GREEN allele recessive and call the BROWN allele dominant, what we're really saying is, "When the alleles GREEN and BROWN are both present in the same person, the expected phenotype from BROWN occurs." This doesn't mean that BROWN keeps the GREEN version from expressing, even - it just means that for one reason or another, BROWN's outward effect (at least, the one we're measuring) "wins" and is visible to us.

It's entirely possible that GREEN is expressing at the gene level, and for some reason it can't yield a phenotype when BROWN is also expressing. Maybe GREEN makes a broken protein; maybe it doesn't yield any protein at all.

...and, as Ziggurat correctly points out, this is all a matter of degrees anyway. To pick a real-world example I do know, familial hypercholesterolemia can come from having one or two defective versions of the cholesterol sensor gene. In the sense that the defective gene leads to detectable disease, it is somewhat "dominant" over the working one. Of course, we could also measure "proper cholesterol regulation" as our phenotype, in which case maybe we might think of the working gene as somewhat dominant. In reality, it's best to think of these things in terms of their actual function - in this case, we just realize that people with two normal sensors have the right dose of regulation, people with one broken gene have a half dose of regulation, and people with two broken genes have no regulation (and won't live all that long without medical intervention).

 

[Amapola]

I don't know if this is outdated... the way I learned about dominant/recessive is that actually there are 4 choices.

An allele can be dominant, recessive, co-dominant or incompletely dominant. Also many traits are controlled by more than one pair of alleles. The more alleles you have that control a trait, the more "exciting" things become, and instead of using say a Punnett square to figure out a simple dominant/recessive trait, one has to resort to binomial equations. This would show a typical bell curve for the inheritance of a trait (for example, how red the kernels of an ear of corn turn out to be). Then of course there are other things in the equation, such as linked traits (traits that are generally found on the same chromosome and are often inherited together) and as has been mentioned, cross-over.

It seems to me really going out on a limb to claim that 50% of the alleles inherited by one sibling are shared by another... that might be true but it might not, as already noted.

I'm more familiar with this in terms of color inheritance in animals. Those are traits that are easily seen by the eye... you can tell easily, for example, what color a rabbit is. It's harder to determine other things, like for example blood type - and blood type in humans has an example of a co-dominant allele. You can't see that just by looking. (Here's a wiki article about dominance and part way down the page it talks about co-dominance.)

 

[jmcvann]

Thanks everyone. Lots of good info here. Seems I was mostly right in what I thought I knew, but the added details help to fill in the blanks.

One more: Wouldn't some recessive traits disappear? Wouldn't a whole lot of green-eyed people have to constantly find each other to keep green eyes going? (To use a possibly weak example.) Is it possible that - using this example - there were lots more eye colors in the past, but all those genes just died out?

Thanks for putting up with my questions. I appreciate all the smarts here at JREF.

 

[sphenisc]

I'm a bit rusty but here a simple rundown. The genetic "sharing" when reproducing is via Chromosomes and a multitude of complex encodings of the genes.

You have 23 pairs of chromosomes(46 total) with 23 of your dad's genes(including Y) and 23 of your mom's Chromosomes (including X). However, the pairs of Chromosomes from your dad are not shared as per a mirror image from your siblings.

For example you have a 50% chance of getting Chromosome 1a or 1b while your brother of sister has a 50% chance of getting either 1a or 1b either. The same with 2a and 2b. For you could have 1a2a while you sister could have 1b2b or 1a2b etc.

So do the math. The chances of you sharing the same sets of chromosomes between yourself and your siblings is definitely not fifty-fifty although it would average out in that way, in fact there is a small chance you could share no genetic similarities between yourself and your sister(remember you and your brother always share a Y chromosome) at all since your sister could receive Chromosomes from both mom and dad that you did not get. About 2*22...very small chance.

You share 50% of your mom and your dad's Chromosome(with the exception for some minor encoding differences) but amount of genetic similarities between yourself and your siblings are based on chance but on average should be about 50%(unless you count in the sex chromosomes)

Anyone else with more recent genetic knowledge please chime in.

At which point your sister says "Yeh, but you share 95% of your DNA with a chimp!"... because sisters are like that.

 

[Ziggurat]

One more: Wouldn't some recessive traits disappear?

Not necessarily. Genes (or more properly, alleles) don't need to be expressed to be passed down. In fact, in some cases it can be easier for an allele to survive if it's recessive. For example, genetic defects which are crippling but recessive can survive because not everyone who has the defective allele is crippled by it. But a dominant defect will generally get wiped out quickly, and only reappear as a mutation.

Is it possible that - using this example - there were lots more eye colors in the past, but all those genes just died out?

Sure, alleles and their corresponding phenotypes can disappear over time. But that can happen regardless of dominance or recessiveness.

 

[Amapola]

Thanks everyone. Lots of good info here. Seems I was mostly right in what I thought I knew, but the added details help to fill in the blanks.

One more: Wouldn't some recessive traits disappear? Wouldn't a whole lot of green-eyed people have to constantly find each other to keep green eyes going? (To use a possibly weak example.) Is it possible that - using this example - there were lots more eye colors in the past, but all those genes just died out?

Thanks for putting up with my questions. I appreciate all the smarts here at JREF.

Well no... what happens is more like this: You start out with one eye color, say brown, and the genes for that are dominant. But then at some point, one of those genes changes... the new allele is sometimes, but not always, recessive. That means that yes, another person has to be carrying that new allele, for that trait to show up (IF it's recessive, some of these are dominant, co-dominant and incompletely dominant). Over time, other alleles appear. If they have to do with eye color, more eye colors show up. (If they have to do with something like being born without a working colon, the offspring die and this trait tends to disappear, or at least be pretty rare. But eye color is fairly harmless and the new trait does not affect the organism in a drastic way.) So that means that over time, you'll have MORE eye colors, where you started out with only one.

An allele can be carried "silently" down many generations, and then match up with a mate, and suddenly that trait appears. People will call this a "throwback" and I guess it is in a way... but it's not that an allele is appearing out of nowhere. It was always there, but since it was recessive, you just couldn't *see* it.

That's why domestic rabbits come in such a bewildering array of colors. They all were domesticated from the European wild rabbit, which comes in only one color, and all the genes that control that color are dominant. But over time, new alleles that caused new coat colors appeared, and since humans like certain colors, these traits were selected for. I've seen pedigreed rabbits suddenly express a color that is nowhere on the papers, even if they go back 10 generations.

 

[paximperium]

One more: Wouldn't some recessive traits disappear? Wouldn't a whole lot of green-eyed people have to constantly find each other to keep green eyes going? (To use a possibly weak example.) Is it possible that - using this example - there were lots more eye colors in the past, but all those genes just died out?

That's assuming Recessive traits are not beneficial in some way even if not dominant.

Recent studies have shown that the reason blue and green eyed genes are prominent is because their primary advantage is because it is pretty...ie. they look hot, get laid more often and make more babies. If I remember correctly, blue eyes actually trace its ancestry back to one ancestor although it likely arose multiple times but died out. And since they are located in cooler and less sun shone areas of the world, they are not as detrimental to eye protection where brown eyes would more likely protect against such brighter light.

The same can also be said for genes that have incomplete Dominance such as Sickle Cell or Cystic Fibrosis. If you have the Ss Sickle Cell trait(one Sickle Cell Diseased trait and one normal), even if you don't show the disease, your Red cells are abnormal. This abnormality is very protective against Malaria so Ss trait carriers are actually more likely to survive in Malaria prone areas and pass on the trait even if it recessive and detrimental if you have ss genotype. CF has the same benefit for diarrhea diseases.

PS Link to the Blue Eye Ancestry: http://evolutiondiary.com/2008/02/01/blue-eyed-humans-have-a-single-common-ancestor/

 

[sanguine]

Not necessarily. Genes (or more properly, alleles) don't need to be expressed to be passed down. In fact, in some cases it can be easier for an allele to survive if it's recessive. For example, genetic defects which are crippling but recessive can survive because not everyone who has the defective allele is crippled by it. But a dominant defect will generally get wiped out quickly, and only reappear as a mutation.

Sure, alleles and their corresponding phenotypes can disappear over time. But that can happen regardless of dominance or recessiveness.

Just to reiterate what Ziggurat was getting at here, the idea of "dominance" is a sort of older way of viewing how a gene led to a detectable phenotype. This doesn't have to have any bearing on the gene's ability to be passed on. Very generally, the things that stop genes from being passed on are (1) chance and (2) detrimental effects of that gene. If there is, for example, no energetic or health cost to "having brown eyes while also having the green-eye gene with no visible phenotype", then there's no selection pressure associated with that gene.

...and as Ziggurat pointed out, "recessive" genes that are detrimental can be carried on quite effectively. If, for example, one of your CFTR alleles is defective, you're probably still a healthy person who can have a very long life. But if you're unlucky enough to catch two broken copies of the CFTR gene, you're going to have cystic fibrosis and a much shorter lifespan. If, in contrast, there were a defective CFTR allele that "gummed up" the working CFTR, it would be much less likely to survive in the population, as anyone with even one copy would die young.

I'm harping a little bit on the idea that dominance and recessiveness is "old" because I think it really is. As we've become progressively more able to actually view the molecular genetic basis of phenotype, I think dominant/recessive just confuses people as a concept, since it makes it sound like the genes fight and one "wins."

(But then, I work professionally in organisms that only have one copy of most of their genes, so I don't even have to think about this all that often...)

 

[paximperium]

At which point your sister says "Yeh, but you share 95% of your DNA with a chimp!"... because sisters are like that.

 

[Dr Adequate]

Here is a primer on genetics.