Extended Data Table 2:
Allele-sharing statistics for deep ancestry
| f 4 (X, Mursi; SA, Han) | f 4 (X, Mota; SA, Han) | f 4 (X, Han; SA, Mursi) | f 4 (X, Mota; SA, Mursi) | |||||
|
| ||||||||
| Test pop | Value | Z -score | Value | Z -score | Value | Z -score | Value | Z -score |
|
| ||||||||
| Dinka | 1.4 | 5.8 | −2.0 | −5.5 | 0.1 | 0.2 | −6.3 | −20.2 |
| Mota | 3.4 | 9.0 | 0 | 0 | 6.3 | 18.1 | 0 | 0 |
| Hadza | 4.1 | 10.3 | 0.8 | 1.7 | 7.3 | 21.2 | 1.0 | 2.7 |
| Yoruba | 4.7 | 17.8 | 1.3 | 3.8 | 5.2 | 18.2 | −1.1 | −3.5 |
| Lemande | 5.0 | 16.8 | 1.7 | 4.5 | 5.7 | 18.2 | −0.6 | −2.1 |
| Mende | 5.7 | 19.1 | 2.3 | 6.3 | 6.3 | 20.0 | 0 | 0 |
| Shum Laka | 11.7 | 38.7 | 8.3 | 22.6 | 12.7 | 40.8 | 6.4 | 20.5 |
| Aka | 13.3 | 39.1 | 9.9 | 25.2 | 13.6 | 40.4 | 7.3 | 22.0 |
| Mbuti | 16.4 | 50.4 | 13.0 | 34.9 | 16.4 | 49.9 | 10.0 | 31.8 |
| Mursi | 0 | 0 | −3.4 | −9.0 | .. | .. | .. | .. |
| Agaw | .. | .. | .. | .. | 0.1 | 0.3 | −6.2 | −18.9 |
| SA | .. | .. | .. | .. | .. | .. | .. | .. |
|
| ||||||||
| f 4 (X, Mursi; SA, Mota) | f 4 (X, Han; SA, Mota) | f 4 (X, Han; SA, Yor) | f 4 (X, Mursi; Chimp, Yor) | |||||
|
| ||||||||
| Test pop | Value | Z -score | Value | Z -score | Value | Z -score | Value | Z -score |
|
| ||||||||
| Dinka | 0.8 | 3.3 | 3.7 | 11.9 | −0.7 | −2.8 | −0.9 | −4.7 |
| Mota | .. | .. | .. | .. | 5.7 | 18.1 | 5.2 | 17.7 |
| Hadza | 4.1 | 11.5 | 7.0 | 17.7 | 4.8 | 15.2 | 3.4 | 11.4 |
| Yoruba | 4.1 | 15.7 | 7.1 | 21.6 | .. | .. | .. | .. |
| Lemande | 4.1 | 14.5 | 7.1 | 21.0 | .. | .. | .. | .. |
| Mende | 4.8 | 17.3 | 7.8 | 22.5 | .. | .. | .. | .. |
| Shum Laka | 9.1 | 29.8 | 12.0 | 33.7 | 8.0 | 28.7 | 8.3 | 31.9 |
| Aka | 10.3 | 33.4 | 13.2 | 35.5 | 7.8 | 24.8 | 8.5 | 30.1 |
| Mbuti | 12.5 | 41.8 | 15.5 | 44.1 | 11.6 | 40.8 | 11.8 | 46.3 |
| Mursi | 0 | 0 | 3.0 | 8.8 | 0.6 | 2.2 | 0 | 0 |
| Agaw | −2.4 | −7.7 | 0.6 | 1.8 | 0 | 0.2 | −0.2 | −0.9 |
| SA | .. | .. | .. | .. | .. | .. | 20.3 | 66.0 |
Variations of allele-sharing statistics (multiplied by 1000; computed on 1,121,119 SNPs) sensitive to ancestry in the test population X from a deeply-splitting lineage, along with Z-scores for difference from zero. We note that the zero level has a different meaning depending on which population is in the second position in the statistic. Blank entries are statistics that are confounded by specific relationships between the test population and one of the reference populations (in the third or fourth position; either duplication of the same group, Agaw with Han due to non-African-related ancestry, or Yoruba with other West Afrians). From the statistics f4 (Mursi/Agaw, Han; South Africa HG, Yoruba), we find minimal differences in deep ancestry proportions among Han, Mursi, and Agaw; from f4 (X, Mursi; Chimp, Yoruba), we obtain a value for South African hunter-gatherers that is roughly twice as large as for Central African hunter-gatherers. SA, ancient South African hunter-gatherers; Yor, Yoruba.