2d, right-hand section). In all simulations, HDAC inhibitor negligible admixture is detected in the Control population. Overall, we have good power to detect 10% admixture that took place 6 Kya, and some power to detect 5% ancient admixture. We genotyped the available Ecuadorian samples at ∼2.5 M sites. The quality of the DNA was low, so it was

necessary to perform whole-genome amplification, and even after this step only about half (16/31) of the samples passed QC. For comparison, we analyzed 11 whole-genome amplified JPT samples in parallel. In order to assess the results, we first compared the genotypes with those of the HGDP populations, using either the worldwide set, or a subset focussed on the relevant populations. Worldwide (Supplementary Fig. 2) and focused Everolimus purchase PCA ( Fig. 3A) both showed that the amplified

JPT fell among the HGDP Japanese, indicating that the amplification procedure and different genotyping chip and centre had no effect detectable by this analysis. Similarly, the Ecuadorians grouped with other Native American populations. ADMIXTURE analyses supported these findings, although with the Ecuadorians tending to form their own cluster at the optimal value of K (Supplementary Fig. 3; Fig. 3B). Some Ecuadorian individuals showed evidence of ancestral components shared with other Native American populations, visible, for example, as the mid green and light green components in Fig. 3B. In addition, in the focussed analysis, two Ecuadorians showed around 5% of a pink ancestral component most prevalent in the Yakut. This component was also detectable at a low level in some Colombians and all of the Maya, as well as in the Russians, so may represent widespread ancient shared ancestry. None of the Ecuadorians showed any of the red component characteristic of the Japanese. This red component was, however, detectable in most of the Maya. While PCA and ADMIXTURE provide a useful visualization of the data, we also performed more formal Reverse transcriptase tests for admixture. TREEMIX again grouped

the Ecuadorians with other Native American populations (Fig. 3C), and when migration was included in the model, the only migration events supported in the focussed group of populations were two events in the Maya (Fig. 3D). We then ran the three-population test and ALDER analysis with all possible population combinations, using Ecuador as a target. No significant results were obtained for either of these two analyses (Table 1), showing that there is no support for migration into the Ecuadorian population. We set out to test whether or not the haplogroup C3* Y chromosomes found at a mean frequency of 17% in two Ecuadorian populations [10] could have been introduced by migration from East Asia, where this haplogroup is common.