xml files for BEAST1.8.0

A total of 370 whole mitochondrial genomes (WMGs) corresponding to Native Central and South American populations were collected from published studies. Of note, the large dataset of 215 Mexican WMGs from Kumar et al. 2011 was not used to prevent an over-representation of Mexican samples. We randomly selected three replicate datasets of 87 sequences, all of which contained exclusively the five Native American haplogroups A2, B2, C1, D1, and D4h3a (table S3). We added 20 East Siberian individuals from sister lineages within haplogroups A, B, C and D in order to estimate the time of the split between Siberian and American lineages. We also added the 92 ancient WMGs from the study, as well as the Anzick-1 mitogenome, to obtain three replicate datasets of 200 sequences. The Anzick-1 WMG was included as it added a deep time calibration at 12.6 kya, was the only ancient D4h3a sequence available, and because the Anzick-1 individual can reasonably be considered a descendant of the first wave of colonizers.