Hello There, Guest!

# How many drugs have race-specific instructions?

(2014-Nov-25, 09:52:12)Emil Wrote: Can you explain the table in more details? I don't quite understand what the data represent.

The point is that when you decompose intra-national racial group outcome differences within and between populations only a "tiny" fraction of the variance is between. Just take a mean difference and transform it using the law of total variance.
(2014-Nov-25, 11:31:04)Zoidberg Wrote: Whats the actual genetic difference between races? Its a frequency difference in some percentage out of the 0.1%. I got 0.014% from Wade, someone else said it was 0.008%(critic of Wade). What is it? Whatever it is you still need to subtract any genes that don't do anything and different genes that do the same thing... then environment. Does anyone have a reasonable estimate of how much of a genetic difference there could be considering everything?

This sounds like a nonsense question. It's usually said that between 10 and 15% of the total within species genetic variance is between continental or regional divisions. You can play all sorts of games with the numbers, though. If you want to make them seem higher you can subtract out intra-individual variance or show that given this difference same continental race individuals are about as related as half sibs relative to the global population. Of you could point to haploid loci which will generally show higher between group variance. If you want to make them look smaller you can focus on the apparently small between group percent and brush aside the fact that in the social sciences a between group difference of 10 to 15% would typically be said to be of moderate to large effect size, and in population genetics to represent a moderate to great difference. Or you can just pick some loci which are fast mutating and so which will show high within group genetic variability and thus lower between. You can also, as you noted, note that humans only differ in coding and non-coding DNA by x percent and set the 10 to 15% or whatever as a fraction of this. In section IV-J of my never complected nature of race paper, I explained why these deflationist arguments are idiotic. Generally, it's futile to try to deduce magnitudes of genetically mediated differences by looking at average genetic variability. And if one grants the logic of the argument one grants virtually all plausible hereditarian hypotheses.
"Whats the actual genetic difference between races?"

We don't know. If we look at most genetic markers (blood groups, enzymes, etc.), genetic differences within human races greatly exceed genetic differences between human races. But we see the same genetic overlap even between many species that are nonetheless anatomically and behaviorally distinct.

Lewontin was trying to compare apples and oranges. Genetic variation within a population differs qualitatively from genetic variation between populations. The first kind of variation is of the sort that cannot be ironed out by similar selection pressures. So we're mostly looking at genes of little or no selective value. The second kind of variation occurs across population boundaries, which tend to be boundaries between different ecosystems, different vegetation zones, different ways of life ... and different selection pressures. So we're looking at genes of greater selective value.

As for the original question of this thread, race-based medicine will continue to be necessary, assuming one cares more about the patient than about ideology. The alternative, individual genotyping, suffers from two serious shortcomings:

1. It's simply too expensive, especially for poor countries or countries with a socialized medical system.
2. In most cases, we still don't know how individual genes relate to specific medical conditions. It's one thing to show that a particular condition has a heritable basis. It's quite another to locate the causal genes.
Both of these points will soon be obsolete for rich countries, as will the second for poor countries.
(2014-Nov-26, 03:49:26)Emil Wrote: Both of these points will soon be obsolete for rich countries, as will the second for poor countries.

Depends on how poor the poor countries will be. There is 3d printing now, and they say it will make personalized medicine much cheaper.

"But we see the same genetic overlap even between many species that are nonetheless anatomically and behaviorally distinct."

Its not the same genetic overlap though because the genes are different. Pointless comparing other species. Only some of the genetic differences between human groups matter. I asked if anyone knows and they don't so thats that. Things like FST between dogs and chimpanzees who are apparently 3 times more diverse even to groups across the same river than humans are across continents is irrelevant.

Only thing that matters are genes that actually do something(worthwhile) different that are then different significantly across these races. If that number is too small then your race medicine argument becomes much less likely to be relevant and thus a waste of time.
Emil,

"Soon" is a slippery word. In the 1970s, most people thought we would soon have colonies on the moon. Didn't happen.

Zoidberg,

"Its not the same genetic overlap"

That was my point. Genetic variation across a population boundary is qualitatively different from genetic variation within a population. The two are not the same and are not comparable.

Many sibling species cannot be reliably distinguished by most genetic markers, even though they can easily be distinguished on the basis of anatomical and behavioral criteria. This is a common finding, and there is no reason to believe that human races are any different.

"Only thing that matters are genes that actually do something"

Again, that was my point. "Functional" genetic variation is much more likely to occur across a population boundary than within a population. Functional variation usually results from differences in selection pressure, such as exists between different populations, whereas non-functional variation usually occurs at genes of little or no selective value and can occur anywhere.
(2014-Nov-26, 18:12:27)Peter Frost Wrote: Emil,

"Soon" is a slippery word. In the 1970s, most people thought we would soon have colonies on the moon. Didn't happen.

Zoidberg,

"Its not the same genetic overlap"

That was my point. Genetic variation across a population boundary is qualitatively different from genetic variation within a population. The two are not the same and are not comparable.

Many sibling species cannot be reliably distinguished by most genetic markers, even though they can easily be distinguished on the basis of anatomical and behavioral criteria. This is a common finding, and there is no reason to believe that human races are any different.

"Only thing that matters are genes that actually do something"

Again, that was my point. "Functional" genetic variation is much more likely to occur across a population boundary than within a population. Functional variation usually results from differences in selection pressure, such as exists between different populations, whereas non-functional variation usually occurs at genes of little or no selective value and can occur anywhere.

Wait between human populations is one overlap, it is comparing the same genes between as within. There is no special difference in the genetic code. Its not like comparing overlap between different species that was my point. You are now saying you cant compare between human populations the same way as within because its like comparing between different species which is illogical.

"Functional" genetic variation is much more likely to occur across a population boundary than within a population"

Proof of this statement above? It seems like utter nonsense.

Non functional variation is seemingly overwhelming in any population. It follows gene flow and even if a functional gene is selected non functional genes will tag along with it. Selection does not work at one gene. People being selected for one can have millions of variations in non functional DNA that will also be selected. It also depends on the selection and then depends on what the gene actually does and if the other variation cant do that via environment.

Can you show me which species show similar overlap as humans? I cant find any. Then we can check the disease differences and how the medications work for them if there are any.
Perhaps one can find information regarding disease treatment in different breeds of livestock (cows, pigs, chickens) and pets (dogs, cats, horses). Dogs have about twice as high F_st value as humans, and I would be surprised if we knew nothing about how medicine interacts with dog breeds.
(2014-Nov-27, 09:03:14)Zoidberg Wrote: Can you show me which species show similar overlap as humans? I cant find any.

Here's a figure from Heller and Siegismund (2009) showing microsatellite genetic differentiation values by heterozygosity for a large sample of species. I added a red line to indicate the typical Human microsatellite Gst. There's a large cluster of species at the low end. I listed some values for species with formally recognizes subspecies here, too.

As for medicine, we can turn to e.g., dog breeds. You run into a drug x breed interaction there too. I imagine that the between population genetic variability is comparable to that between human wild breeds.