Most modern Arabians can be modeled as deriving all of their Levantine/Anatolian-related ancestry from a Neolithic farmer-related source; however, the medieval Soqotri had relatively more ancestry from ancient groups closer to Natufian hunter-gatherers (2/n).

We observe this same pattern in only one other group - present-day people from the Hadramawt region of present-day Yemen - which represents a likely source for the peopling of Soqotra. Genetic connections between these regions are extremely strong.(3/n)

The medieval Soqotri gene pool is best modeled as deriving ~86% ancestry from people from the Hadramawt, with the remaining ancestry well-proxied by an Iranian related source with up to 2% ancestry from the Indian sub-continent. (4/n)

This is consistent with the location of Soqotra along Indian Ocean trade routes: Indian inscriptions in Hoq Cave are written in varieties of Brāhmī script and in Sanskrit or vernacularized Sanskrit & the Periplus records India’s provision of female slaves to Soqotra (5/n)

However, traders and mariners (from India and elsewhere) rarely moved from coastal areas and plausibly not having much of a genetic impact, which we document in the stability of the Soqotri gene pool over the span of more than a millennium. (6/n)

Despite being located off the Horn of Africa, medieval Soqotri have extremely little sub-Saharan African-related ancestry: models of genome-wide data can be fit without this ancestry, while we find evidence of only one L3 mtDNA haplogroup. (7/n)

With ROH, we see a reduced rate of cousin marriages in medieval relative to present-day Soqotra. More research will show if there was a relatively recent shift in marriage practices btwn medieval times and today on Soqotra and perhaps in other parts of the Arab world. (8/n)

A longstanding question in Soqotri archaeology asks if burial tafoni were segregated by sex? We show that 7/13 tafoni included individuals of both sexes & provide evidence of matrilineal & patrilineal relationships among people buried together. (9/n)

This National Geographic project would not have happened without the amazing men & women on Soqotra who contributed to this collaboration. Thank you to GOAM especially for permitting this study. I’m incredibly proud of the community engagement aspect of this work. (10/n)

A huge thank you to Nada Salem for translating our abstract into Arabic, which helped us share results with people on Soqotra (more of this to come!) (11/n)