Table 1.
Population groupings tested | P-value for this many Asian streams being enough to explain the data |
Minimum number of streams of Asian gene flow needed to explain the data |
||
---|---|---|---|---|
1 | 2 | 3 | ||
E. Greenland Inuit / W. Greenland Inuit / First American | <10−9 | 0.64 | 1 | 2 |
E. Greenland Inuit / Aleutian / First American | <10−9 | 0.57 | 1 | 2 |
W. Greenland Inuit / Aleutian / First American | <10−9 | 0.41 | 1 | 2 |
Chipewyan / E. Greenland Inuit / First American | <10−9 | 0.02 | 1 | 3 |
Chipewyan / W. Greenland Inuit / First American | <10−9 | 0.006 | 1 | 3 |
Chipewyan / Aleutian / First American | <10−9 | 0.03 | 1 | 3 |
Saqqaq / E. Greenland Inuit / First American | <10−9 | 6×10−6 | 1 | 3 |
Saqqaq / W. Greenland Inuit / First American | <10−9 | 2×10−6 | 1 | 3 |
Saqqaq / Aleutian / First American | <10−9 | 0.17 | 1 | 2 |
Saqqaq / Chipewyan / First American | <10−9 | 0.29 | 1 | 2 |
Saqqaq / Eskimo-Aleut / Chipewyan / First American | <10−9 | 8×10−6 | 0.27 | 3 |
Notes: We use the method described in Note S6 to test formally whether specified groupings of Native American populations are consistent with descending from 1, 2, or 3 streams of gene flow from Asia. We use “First American” to refer to a pool of 43 populations from Meso-America southward, and “Eskimo-Aleut” to refer to a pool of East and West Greenland Inuit and Aleuts. We test either 3 or 4 population groupings (when there are 3 groupings, the maximum number of streams we can reject is 2, and so the P-value for 3 streams is always 1). At least two streams of Asian gene flow are required to explain all rows (P<10−9). The Chipewyan, Eskimo-Aleut and First Americans can only be jointly explained by at least three streams. Analysis of the Saqqaq Paleo-Eskimo (using ~6-fold fewer SNPs than for the other analyses) show that the Asian ancestry in this individual has a component that is different from that in First Americans and Greenland Inuit, but indistinguishable from the Chipewyan.