Looks like my IntenseDebate comments system is hosed. At least it doesn't work consistently.
So, to reply to a comment until I figure out how to make things work:
Suppose you said that X and Y had a similar ancestry profile of a Norwegian and Italian mix, except that Y also had 22% Japanese ancestry. It makes one wonder, what does a "similar ancestry profile" mean?
How about "X is half tin and half copper" vs. "Y consists of equal parts of tin and copper with 22% arsenic"? Seems like the obvious analogy to me.
---
This analogy doesn't work for me for several reasons:
First of all, X and Y are different alloys, some variety of bronze versus arsenical bronze.
Next: the X itself consists of 14-42% AASI (and the rest Iranian farmer and Siberian Hunter Gatherer)
So this is like saying X is 14% tin and 86% copper or 42% tin and 58% copper and Y is various proportions of tin, copper and arsenic (arsenic averaging 22%).
Next: What does "similar ancestry profile" mean of X and Y when X and Y don't share at least 22% of their ancestors?
Next: since when is 78% of a genome similar to 100% of another?
Next: A majority of Indians are somewhere on the ANI/ASI cline with little other admixture. Would you say they "share the same ancestry profile"?
Next: Out of 246 groups of Indians that this pre-print had genetic profiles for, they excluded 106. Some for paucity of data; but others for having something more than just ANI/ASI. i.e., for that part of the analysis, these 106 groups did not have "similar ancestry profiles".
Next: just as a practical thing - if someone said, I have the same ancestry profile as European Americans but also 22% African - well, unless they are "passing" this someone has gone from being white to being black. Trump supporters (well, some of them) would likely riot over a statement that they share an ancestry profile.
---
IMO, on an entirely different tack, I get a suspicion that some of the genetics people don't really understand the algorithms that underlie their computational analytic apparatus; it is largely black-boxes to them.
---
PS: Let's get to some amusing applications of "similarity".
These two geometric figures are similar though one is triangle and the other is a quadrilateral.
"I'll offer you a similar deal as I offered him, except you'll pay 22% more."
So, to reply to a comment until I figure out how to make things work:
Suppose you said that X and Y had a similar ancestry profile of a Norwegian and Italian mix, except that Y also had 22% Japanese ancestry. It makes one wonder, what does a "similar ancestry profile" mean?
How about "X is half tin and half copper" vs. "Y consists of equal parts of tin and copper with 22% arsenic"? Seems like the obvious analogy to me.
---
This analogy doesn't work for me for several reasons:
First of all, X and Y are different alloys, some variety of bronze versus arsenical bronze.
Next: the X itself consists of 14-42% AASI (and the rest Iranian farmer and Siberian Hunter Gatherer)
So this is like saying X is 14% tin and 86% copper or 42% tin and 58% copper and Y is various proportions of tin, copper and arsenic (arsenic averaging 22%).
Next: What does "similar ancestry profile" mean of X and Y when X and Y don't share at least 22% of their ancestors?
Next: since when is 78% of a genome similar to 100% of another?
Next: A majority of Indians are somewhere on the ANI/ASI cline with little other admixture. Would you say they "share the same ancestry profile"?
Next: Out of 246 groups of Indians that this pre-print had genetic profiles for, they excluded 106. Some for paucity of data; but others for having something more than just ANI/ASI. i.e., for that part of the analysis, these 106 groups did not have "similar ancestry profiles".
Next: just as a practical thing - if someone said, I have the same ancestry profile as European Americans but also 22% African - well, unless they are "passing" this someone has gone from being white to being black. Trump supporters (well, some of them) would likely riot over a statement that they share an ancestry profile.
---
IMO, on an entirely different tack, I get a suspicion that some of the genetics people don't really understand the algorithms that underlie their computational analytic apparatus; it is largely black-boxes to them.
---
PS: Let's get to some amusing applications of "similarity".
These two geometric figures are similar though one is triangle and the other is a quadrilateral.
"I'll offer you a similar deal as I offered him, except you'll pay 22% more."
No comments:
Post a Comment