3 ms·

The Rosenberg et al paper just says that at a particular K partition (5 in this case) you get genetic structures that closely map onto what Americans view as r

The Rosenberg et al paper just says that at a particular K partition (5 in this case) you get genetic structures that closely map onto what Americans view as race. It also says that at K=1 there are only Africans,and at k=3 there are only Africans, Europeans and Asians. If you’re gonna say these are races, why stop at k=5? Why not look at what STRUCTURE pops out for k= 25?

bobcostas55 9y ago
It's absolutely true that there are genetic clusters on multiple levels, and the choice of K is to a certain extent arbitrary. But the fact that there are multiple ways of slicing things up does not affect the reality of the underlying structure. The use of "race" to denote 25 different groups would probably be a bit confusing though, unfortunately we don't have any better terms ("genetic cluster" is a bit awkward?) Here's a really cool chart from an ancient DNA paper with stacked charts from K=2 to K=20: https://www.ncbi.nlm.nih.gov/core/lw/2.0/html/tileshop_pmc/tileshop_pmc_inline.html?title=Click%20on%20image%20to%20zoom&p=PMC3&id=4170574_nihms613260f5.jpg https://www.ncbi.nlm.nih.gov/core/lw/2.0/html/tileshop_pmc/t... (The paper: https://www.biorxiv.org/content/early/2013/12/23/001552 https://www.biorxiv.org/content/early/2013/12/23/001552) Zoom in far enough and the clusters basically replicate national borders: http://scienceblogs.com/notrocketscience/2008/09/01/european-genes-mirror-european-geography/ http://scienceblogs.com/notrocketscience/2008/09/01/european... so you might use the term "ethnicity" for them.
- novacole 9y ago
  My Masters thesis covered a lot of what is in these papers. Here are a few possibly surprising things people who think genetic clusters point to races may not realize: 1) People from Papua New Guinea and Chinese people would be considered the same race according to genetic clusters at K=4. 2) At k=2 there are only Africans and Asian races. 3) And Europeans and dravidians are part of the same race, at k=5.

bobcostas55 9y ago

It's absolutely true that there are genetic clusters on multiple levels, and the choice of K is to a certain extent arbitrary. But the fact that there are multiple ways of slicing things up does not affect the reality of the underlying structure. The use of "race" to denote 25 different groups would probably be a bit confusing though, unfortunately we don't have any better terms ("genetic cluster" is a bit awkward?) Here's a really cool chart from an ancient DNA paper with stacked charts from K=2 to K=20: https://www.ncbi.nlm.nih.gov/core/lw/2.0/html/tileshop_pmc/tileshop_pmc_inline.html?title=Click%20on%20image%20to%20zoom&p=PMC3&id=4170574_nihms613260f5.jpg https://www.ncbi.nlm.nih.gov/core/lw/2.0/html/tileshop_pmc/t... (The paper: https://www.biorxiv.org/content/early/2013/12/23/001552 https://www.biorxiv.org/content/early/2013/12/23/001552) Zoom in far enough and the clusters basically replicate national borders: http://scienceblogs.com/notrocketscience/2008/09/01/european-genes-mirror-european-geography/ http://scienceblogs.com/notrocketscience/2008/09/01/european... so you might use the term "ethnicity" for them.

novacole 9y ago
My Masters thesis covered a lot of what is in these papers. Here are a few possibly surprising things people who think genetic clusters point to races may not realize: 1) People from Papua New Guinea and Chinese people would be considered the same race according to genetic clusters at K=4. 2) At k=2 there are only Africans and Asian races. 3) And Europeans and dravidians are part of the same race, at k=5.